Few statutes in American civil rights law carry a legislative history that reads like an answer key. The ADA Amendments Act of 2008 is one of them. Congress did not merely change the definition of disability; it named the Supreme Court decisions it was reversing, quoted the reasoning it was rejecting, and wrote the replacement rules into the United States Code with a level of explicitness that leaves little room for the usual arguments about what the legislature intended. Public Law 110-325 was an override, and it behaved like one. It took a definition that courts had read narrowly for nine years and rebuilt it so that the reading could not be repeated, while leaving every duty, every defense, and every remedy exactly where the 1990 statute had put them. The result was not a new obligation for employers, governments, or businesses. It was a relocation of the fight, from the threshold question of who counts as disabled to the practical questions of who is qualified and what change is reasonable.

The story matters beyond disability law because it is a case study in how the American system corrects judicial narrowing without constitutional confrontation. Congress did not attack the Court. It did not strip jurisdiction or denounce the justices. It wrote a better definition, with findings that explained the disagreement and purposes that directed the new reading, and it passed the result with votes so lopsided that opposition was a rounding error. The override cycle, judicial narrowing followed by legislative restoration, is one of the recurring rhythms of civil rights legislation, and the 2008 act is among its cleanest examples. This guide follows that rhythm from the 1990 promise through the narrowing years to the restoration, keeping the statute’s text at the center throughout.

The definition Congress wrote in 1990

The original Americans with Disabilities Act defined disability in three prongs, and the three prongs matter because the 2008 revisions touched each of them differently. A person counted as disabled under the 1990 law if she had a physical or mental impairment that substantially limited one or more major life activities, if she had a record of such an impairment, or if she was regarded as having such an impairment. The first prong did the most work. It required an impairment, a substantial limitation, and a major life activity, and each of those terms became a separate battlefield. The second prong protected people with histories of serious conditions, such as a cancer survivor in remission, whose past impairment still shadowed hiring decisions. The third prong reached a different wrong altogether: discrimination based on what others believed about a person, whether or not the belief was medically accurate. The guide to the 1990 statute sets out the five titles and the full enforcement map; this article stays with the definition, because the definition is what the courts narrowed and what Congress restored.

The three-pronged design was itself an act of legislative foresight. Congress understood that disability discrimination takes multiple forms: exclusion based on actual limitation, exclusion based on history, and exclusion based on stereotype. A single-pronged definition would have missed the second and third forms, leaving the cancer survivor and the misperceived worker without recourse. The prongs overlap by design, so that a plaintiff who fails one may satisfy another. That redundancy became critical after the trilogy, when the first prong’s narrowing pushed plaintiffs toward the third, only to find the third narrowed as well. The 2008 act’s prong-by-prong approach to repair reflects an understanding of the original design: each prong does distinct work, and each needed its own fix.

Congress had written the 1990 findings to signal breadth. The findings in 42 U.S.C. 12101 described discrimination against people with disabilities as a serious and pervasive social problem and stated a national goal of assuring equality of opportunity, full participation, independent living, and economic self-sufficiency. Nothing in that language suggested a cramped reading of the protected class. The definition itself used ordinary words, and ordinary words invite judges to supply their meaning from context, precedent, and policy instinct. Over the following decade, the Supreme Court supplied meanings that steadily shrank the class. By the early 2000s, employment lawyers on both sides understood that the hardest part of a disability discrimination case was often not proving the employer acted badly. It was proving the plaintiff was disabled at all. That inversion, in which the threshold question consumed the case, is the problem the 2008 act was written to end.

Why three prongs

The three-pronged structure was itself a legislative choice worth understanding, because the 2008 act treats each prong differently. The first prong, actual disability, covers people whose impairments presently and substantially limit them. It is the prong most cases are litigated under, and it received the four mechanical changes. The second prong, record of impairment, covers people whose past conditions still shape how others treat them: the cancer survivor, the formerly hospitalized worker, the person with a history that follows her through background checks and office gossip. The 2008 act left the record-of prong largely alone, because the courts had not narrowed it the way they narrowed the first prong, and because the episodic-or-in-remission rule now lets many of these plaintiffs proceed under the first prong directly.

The third prong, regarded as, covers people who may have no limiting impairment at all but face discrimination based on what others believe. It is the prong the 2008 act restructured most dramatically, trading a broadened test for two new boundaries. The three prongs thus tell the story of the amendments in miniature: the first prong rebuilt, the second prong confirmed, the third prong rebalanced. A reader who can say which prong a given fact pattern belongs in, and which 2008 change affects it, has mastered the operational core of the act.

Why did Congress say the courts had narrowed the statute?

Congress said so directly in the findings, which is unusual. Section 2(a)(4) of the 2008 act names the Sutton trilogy and Toyota v. Williams and states that their holdings narrowed the broad scope of protection the 1990 law was intended to afford, making the override explicit rather than a matter of inference.

What the 1990 findings promised

The original statute’s findings described disability discrimination as a serious and pervasive social problem and announced a national mandate to end it. The findings spoke of millions of Americans excluded from the mainstream of American life by prejudice, by architectural barriers, and by the failure of employers and public entities to make the changes that would admit them. The remedial purpose was stated in the broadest terms: the statute was to provide a clear and comprehensive national mandate for the elimination of discrimination against individuals with disabilities.

The 2008 Congress read those findings as a promise of breadth and accused the courts of breaking it. The new findings state that the Supreme Court’s decisions narrowed the broad scope of protection the 1990 statute was intended to afford, and the purposes then direct the courts back to the breadth the findings describe. Whether the 1990 Congress truly intended the breadth the 2008 Congress claimed is a question of legislative history that the 2008 findings answer in the affirmative by fiat. But the structure of the argument is what matters for understanding the statute: Congress presented itself as the faithful interpreter of its predecessor’s intent, and the courts as the faithless ones.

The enforcement architecture of the original statute reinforced the promise of breadth. Five titles spread the definition across employment, public services, public accommodations, telecommunications, and miscellaneous provisions, with the EEOC, the Department of Justice, and other agencies assigned enforcement roles. The definition was the common key to all five titles: the same three prongs opened the courthouse door for the employee, the transit rider, and the restaurant patron alike. When the courts narrowed the key, they narrowed it for every title at once, and when Congress restored it in 2008, the restoration flowed through the same five channels.

Why the 1990 words invited narrowing

The 1990 definition’s vulnerability was a product of its virtues. Congress wrote it in plain language, impairment, substantially limits, major life activities, because plain language signals breadth and invites ordinary understanding. But plain language is also an invitation to judges, and judges bring theories. The trilogy and Toyota show what happens when a court approaches open-textured words with a narrowing theory: each term becomes a separate gate, and each gate gets a strict construction. Impairment was read through mitigation. Substantially was read as demanding. Major was read as central to most people’s daily lives. The three readings multiplied, and their product was a protected class far smaller than the findings contemplated.

The mismatch between the findings and the definition is the key to the whole episode. The findings in 42 U.S.C. 12101 speak in the language of broad social reform: pervasive discrimination, national goals, full participation. The definition speaks in three spare prongs. When the Court construed the prongs strictly, it could claim fidelity to the definition’s words while defeating the findings’ ambitions, and no textualist objection could easily answer, because the words really are open to the strict reading. The lesson Congress drew was that findings alone do not protect a definition. The 2008 act therefore pairs its findings with purposes that do interpretive work and with Code provisions that write the broad reading into law. The definition’s words are still plain. The instructions for reading them no longer permit the narrowing theory.

There is a broader point here about how civil rights statutes age. A broad statute enacted with fanfare will always face the question of whether its ambitions survive contact with courts that were not parties to the bargain. The 1990 act’s drafters assumed that the findings’ breadth would carry into the definition’s application. Nine years of case law proved the assumption wrong. The 2008 act’s drafters made no such assumption. They wrote the breadth into the operative provisions, named the decisions that had defeated it, and directed the implementing agency to hold the line. The difference between the two drafting strategies is the difference between hoping for broad construction and commanding it.

What made the 1990 definition vulnerable to narrowing?

The definition used plain, open-textured words that invited strict judicial construction, while the broad ambitions lived only in the findings. Each key term became a separate gate, the gates multiplied, and the protected class shrank far below what the findings contemplated.

Three decisions in one day

On June 22, 1999, the Supreme Court decided three cases that together rewrote the first prong of the disability definition. The trilogy, Sutton v. United Air Lines, Inc., 527 U.S. 471; Murphy v. United Parcel Service, Inc., 527 U.S. 516; and Albertson’s, Inc. v. Kirkingburg, 527 U.S. 555, addressed the same core question from three angles: when a court decides whether an impairment substantially limits a major life activity, should it consider the person as she is with her medication, her devices, and her coping strategies, or as she would be without them.

Sutton involved plaintiffs with severe myopia whose vision was corrected with lenses. The Court held that the determination of whether an individual is disabled should be made with reference to measures that mitigate the individual’s impairment. With corrective lenses, the plaintiffs’ vision did not substantially limit the major life activity of seeing, so they were not disabled under the statute, even though the employer had acted on the uncorrected condition. The opinion treated the use of mitigating measures as part of the disability determination itself.

Murphy applied the same rule to hypertension controlled by medication. The Court held that the plaintiff’s impairment, considered in its medicated state, did not substantially limit a major life activity. The decision confirmed that Sutton was not limited to corrective lenses. Medication counted as mitigation, and a person whose condition was well managed by drugs stood outside the protected class.

Albertson’s extended the principle further than many observers expected. The Court held that mitigating measures include measures undertaken, whether consciously or not, with the body’s own systems. The phrase mattered. Mitigation was no longer confined to pills and devices. It included the body’s unconscious adaptations, the neurological adjustments that a person never chose and could not switch off. Taken together, the three opinions established a rule that the statute’s critics would summarize with bitter precision: the better a person managed a condition, the less protection the law gave her. A fuller account of each decision’s doctrine and current status appears in the companion treatment of the Supreme Court disability cases that narrowed coverage before Congress acted.

The practical consequence was immediate and strange. Employees with epilepsy controlled by medication, diabetes managed with insulin, depression treated with antidepressants, and hypertension kept in check by drugs found themselves in a double bind. Their conditions were serious enough to affect hiring, promotion, and retention decisions, because employers noticed the diagnosis, the medication schedule, or the occasional breakthrough symptom. But the same conditions, viewed through the lens of the mitigating measures, were not serious enough to count as disabilities. The disability community described this as a catch-22: too functional to be protected, too impaired to be hired. The description came from advocates who watched clients lose at the threshold, before any court ever reached the question of whether the employer had behaved unlawfully.

Inside the trilogy’s reasoning

The three opinions of June 22, 1999, deserve a closer reading than the summary above, because Congress’s reversal was aimed at their reasoning as well as their results. Sutton’s core move was textual: the opinion reasoned from the statute’s present-tense phrasing, which speaks of an impairment that substantially limits a major life activity, to the conclusion that the impairment must be assessed in its present, mitigated state. A person whose disability is corrected by mitigating measures, the opinion held, does not have an impairment that presently substantially limits a major life activity. The reasoning treated the definition as a snapshot of the body as it stands, corrected and all, rather than as a judgment about the underlying condition.

Sutton’s treatment of the regarded-as prong was equally consequential. The sisters argued that United had regarded them as disabled by enforcing its uncorrected-vision standard. The Court rejected the claim, holding that the employer’s requirement did not show it regarded them as substantially limited in the major life activity of seeing, but only as unable to satisfy the particular requirements of the global pilot position. The prong that Congress had written as a protection against action based on myth and stereotype became, under this reasoning, a demand for proof about the employer’s internal theory of the impairment’s severity. Few plaintiffs could meet it, because employers rarely announced such theories.

Murphy extended the logic to medication and to the regulatory context. The plaintiff’s hypertension was controlled by drugs, and the employer had acted on the basis of federal Department of Transportation certification standards. The Court applied Sutton’s rule to hold that the medicated condition was not a disability, and it held that the employer’s reliance on the federal standard did not show that it regarded the plaintiff as substantially limited in working. The decision thus insulated two kinds of employer conduct at once: the medical judgment that a medicated worker was not disabled, and the regulatory judgment that an uncertified worker could not drive.

Albertson’s pushed the doctrine to its furthest reach. The plaintiff’s monocular vision had been unconsciously compensated by his own neurological adaptations, and the Court held that those adaptations counted as mitigating measures even though no device, drug, or conscious strategy was involved. Measures undertaken with the body’s own systems, whether consciously or not, entered the disability calculus on the same terms as eyeglasses and medication. The opinion then remanded for an individualized assessment of substantial limitation, a directive that underscored how granular and plaintiff-burdensome the threshold inquiry had become. After Albertson’s, there was no corner of the plaintiff’s physiology the court could not credit against the claim.

Why the trilogy proved so hard to work around

The Sutton rule had a structural feature that made it unusually resistant to case-by-case softening. It did not announce an exception or a balancing test. It defined the order of operations: first account for every mitigating measure, then ask whether what remains substantially limits a major life activity. Once that sequence was fixed, the outcome in most cases was foreordained, because modern medicine is good at its job. Insulin controls diabetes. Anticonvulsants control seizures. Antidepressants lift the symptoms they target. A rule that measures disability after treatment will always find less disability than a rule that measures it before, and the gap grows with every advance in treatment. The trilogy thus built a legal regime whose coverage shrank as medicine improved, an odd result for a civil rights statute.

The rule also created an incentive problem that disability advocates pressed throughout the legislative debate. If treatment removes legal protection, then a worker who manages a condition diligently is worse off, legally, than one who does not. No court held that anyone should forgo treatment to preserve a claim, and the opinions did not encourage that result. But the structure of the rule pointed in that direction, and advocates argued that the law should never make the responsible management of a health condition a litigation disadvantage. The 2008 act answers the concern structurally rather than rhetorically: by directing that mitigating measures be disregarded, it severs the link between good treatment and lost coverage entirely.

Lower courts, faced with the trilogy, applied it the way lower courts apply clear Supreme Court instructions: broadly and literally. The opinions’ logic extended naturally to new contexts, and litigants tested its boundaries with prosthetics, hearing aids, psychotherapy, and learned coping strategies. Each extension followed the same sequence, account for the mitigation, then measure what remains, and each extension confirmed the disability community’s account of the era. The definition was not merely narrow at its edges. It was narrow at its center, because the conditions most people think of as disabilities, epilepsy, diabetes, serious mental illness, are precisely the conditions that modern treatment manages well enough to fail the test.

The trilogy’s narrowing of the third prong compounded the problem. Sutton did not only address mitigating measures under the first prong. Its reasoning also constricted regarded-as coverage, requiring a plaintiff to show that the employer perceived an impairment as substantially limiting. That reading made the third prong nearly useless as a backstop, because an employer who acted on a perceived impairment could defeat the claim by showing that the perceived impairment, too, was mitigated or minor. The 2008 act’s purposes single out this reasoning for rejection in section 2(b)(3), restoring the earlier and broader understanding associated with School Board of Nassau County v. Arline. The override was thus aimed at the trilogy’s full footprint, not only its most famous holding.

The trilogy also shaped the professional culture of disability practice in ways that outlasted the holdings themselves. A generation of defense lawyers built their expertise on threshold motions, and a generation of plaintiffs’ lawyers built theirs on surviving them. Continuing legal education programs taught the trilogy as the organizing framework of the field. Treatises structured their disability chapters around the mitigating-measures rule and the demanding standard. When Congress reversed the holdings, it did not only change the law. It obsoleted an entire professional skill set, the art of winning or surviving the threshold fight, and replaced it with a different one, the art of litigating qualification and accommodation. The speed with which the bar adapted testifies to the clarity of the new text: there was little to argue about, so lawyers moved on to the questions the statute had always meant them to argue.

Why did the trilogy survive so many challenges before Congress acted?

The trilogy stated a clear order of operations that lower courts could apply mechanically, and only Congress or the Supreme Court itself could displace it. Litigants tried narrowing distinctions for nine years, but the rule’s logic extended to each new mitigating measure, so the definition kept shrinking until the legislature intervened.

The compound rule: Sutton plus Toyota

Neither Sutton nor Toyota, standing alone, would have produced the pre-2009 docket. Together they formed a compound rule whose two halves reinforced each other at every step of the analysis. Sutton shrank the impairment by crediting every correction, adaptation, and medication against the plaintiff. Toyota raised the measuring stick by demanding a strict, exacting showing of substantial limitation. A plaintiff therefore had to demonstrate a severe restriction of centrally important daily activities, measured from a baseline that had already been reduced by every ameliorative measure the body or medicine supplied. The doctrines multiplied rather than added: each made the other’s burden heavier.

The lower courts applied the compound rule with the thoroughness the opinions invited. District judges parsed medical records for evidence of mitigation, weighed expert testimony about hypothetical unmedicated severity, and measured the resulting picture against Toyota’s demanding standard. The inquiry was individualized in the way Albertson’s required and exacting in the way Toyota required, which meant it was expensive, slow, and tilted toward the defense. Plaintiffs who could afford the experts and survive the motions sometimes reached a jury; most did not.

The compound rule also produced a distinctive evidentiary paradox. The better the plaintiff’s medical management, the more mitigation there was to credit, and the thinner the remaining limitation. The worse the management, the less mitigation to credit, but the harder it was to satisfy Toyota’s demand for severity. Plaintiffs were squeezed from both directions: successful treatment erased the limitation, and unsuccessful treatment rarely produced the kind of severe restriction the demanding standard required. The doctrines left a narrow band of qualifying plaintiffs, those with serious impairments resistant to correction, and excluded the broad middle of workers with managed conditions that the 1990 findings had described.

Congress’s 2008 findings describe this compound rule as the narrowing of the broad scope of protection the statute was intended to afford. The description is Congress’s, and it is worth noting that the findings name the Sutton line and Toyota together, as a pair. The statute’s answer was correspondingly paired: the mitigating-measures disregard undoes Sutton, and the broad-construction command undoes Toyota. Each half of the compound rule received its own statutory reversal, and the two reversals together dismantle the machine the two doctrines had built.

The demanding standard

Three years later, the Court tightened the definition from the other direction. In Toyota Motor Manufacturing, Kentucky, Inc. v. Williams, 534 U.S. 184 (2002), the question was what “substantially limits” meant for a worker with carpal tunnel syndrome and related impairments who claimed she was limited in the major life activity of performing manual tasks. The Court construed the terms strictly. It held that the words “substantially” and “major” needed to be interpreted strictly to create a demanding standard for qualifying as disabled, and that to be substantially limited in performing manual tasks, a person must have an impairment that prevents or severely restricts her from doing activities that are of central importance to most people’s daily lives.

The opinion did two things at once. It raised the bar for what counted as a substantial limitation, and it narrowed the comparison set: the question was not whether the impairment limited the person relative to her own prior capacity or her job’s demands, but whether it prevented or severely restricted activities central to most people’s daily lives. Lower courts read Toyota as an instruction to demand extensive evidentiary showings at the threshold. Plaintiffs were asked to prove, often with expert testimony, the precise degree to which an impairment restricted daily activities of central importance, and many cases ended on summary judgment before discovery into the employer’s conduct had gone far. The definition had become a demanding standard by judicial construction, and the construction sat in direct tension with the broad remedial purpose Congress had stated in 1990.

What Toyota demanded of plaintiffs

To feel the weight of the Toyota standard, consider what a plaintiff had to assemble before a court would reach the employer’s behavior. She needed medical evidence of the impairment, functional evidence of the limitation, and comparative evidence that the limitation was severe relative to the daily lives of most people. Each element invited a defense motion. The medical evidence could be challenged as insufficiently specific. The functional evidence could be discounted because the plaintiff had adapted, which under Albertson’s counted as mitigation anyway. The comparative evidence required showing that the restricted activities were of central importance to most people’s daily lives, a phrase that generated its own satellite litigation about which activities qualified. A worker limited in lifting, bending, or performing manual tasks might learn that the court did not regard those tasks as central enough, or that her remaining abilities were sufficient to defeat the claim.

The standard also interacted badly with the realities of chronic illness. Many serious conditions fluctuate. A worker might be severely restricted during a flare and functional between episodes, and Toyota’s framework gave courts little guidance on which moment to measure. The 2008 act’s episodic-or-in-remission rule answers that question directly, but before the act, the demanding standard turned fluctuation into a defense argument: the good days proved the limitation was not substantial. The result was a body of threshold law that rewarded employers for emphasizing what a plaintiff could still do and punished plaintiffs for resilience, adaptation, and effective treatment. Congress’s rejection of the demanding standard in section 2(b)(4) was aimed at this entire apparatus, not only at the sentence in Toyota that announced it.

The view from the compliance office

The threshold era looked different from inside a human resources department than it did from a courtroom, and the employer perspective deserves its own statement. For compliance officers, the trilogy and Toyota offered something valuable: a screening device. When a worker requested an accommodation, the first question was whether the worker was disabled at all, and the demanding standard meant the answer was often no. That answer ended the inquiry. No interactive process, no hardship analysis, no documentation of essential functions. The threshold defense was not only a litigation tool. It was a way of managing the daily flow of accommodation requests, and its availability shaped how employers staffed and trained their human resources functions.

The 2008 act took that screening device away, and the compliance consequence was a reallocation of effort from the threshold to the process. Employers who had treated the interactive process as optional discovered that it was now the central compliance obligation. The process requires a structured dialogue: acknowledging the request, gathering medical information sufficient to understand the limitation, identifying potential accommodations, and documenting the decision. Each step generates a record, and the record is what defends the employer if the decision is later challenged. The shift from threshold screening to process documentation raised the fixed costs of compliance, more training, more forms, more careful file-keeping, while lowering the marginal cost of each individual decision, because the definition no longer had to be litigated first.

Employer advocates describe this reallocation with mixed feelings, and the mixed feelings are worth reporting accurately. On one hand, the interactive process is a better way to run a workplace than threshold screening. It surfaces problems early, it often produces cheap solutions, and it generates the documentation that wins cases. Many large employers had already built robust processes before 2009 and experienced the amendments as a ratification of their approach. On the other hand, smaller employers without dedicated human resources staff experienced the change as a new administrative burden, a set of procedures to learn and maintain for a category of requests that the old threshold had often made disappear. The statute does not calibrate its procedural expectations to employer size in the definition, and the compliance costs fall unevenly. That unevenness is the most sympathetic version of the employer concern about broadened coverage, and it survives the restoration framing: even if the duties are the ones Congress always intended, the work of performing them is real.

The compliance story also explains why employer organizations ultimately supported the bill rather than merely acquiescing in it. A clear, broad definition with predictable boundaries is easier to administer than a narrow definition with unpredictable edges. Under the trilogy, every accommodation request began with a judgment call about whether the worker was disabled enough, a call that line managers were ill-equipped to make and that generated inconsistent results across departments. Under the amendments, the threshold question largely answers itself, and managers can focus on the practical work of finding accommodations. Clarity has compliance value even when it expands the protected class, because the cost of uncertainty often exceeds the cost of coverage. The negotiated boundaries, the eyeglasses exception, the transitory-and-minor rule, the regarded-as accommodation limit, gave employers the predictability they needed to accept the clarity advocates wanted.

The years of the threshold fight

Between 1999 and 2008, the structure of disability employment litigation changed shape. Before the trilogy, the central questions in a typical case were whether the plaintiff was qualified for the job and whether the employer had failed to provide a reasonable accommodation. After the trilogy and Toyota, the central question moved to the front of the case: was this person disabled at all. Defense counsel learned that the cheapest way to win was to attack the definition. Plaintiffs’ lawyers learned that they needed medical experts, functional capacity evaluations, and detailed testimony about daily activities before they could reach the merits. The threshold question consumed discovery budgets and produced a body of case law that read less like civil rights doctrine and more like a medical coding manual.

The dynamic fed on itself. Each defense victory at the threshold generated precedent that made the next threshold fight harder for plaintiffs, and each published opinion added another gloss on what substantially limits meant in a new factual context. Trial judges, faced with a demanding standard from the Supreme Court and a growing library of defense-friendly applications, granted summary judgment at rates that reshaped practice. Plaintiffs’ lawyers began screening potential clients not for the strength of the discrimination evidence but for the severity of the medical file. A worker with a vivid story of unfair treatment but a well-managed condition was a bad investment, because the case would die before the story could be told. The statute’s promise of individualized assessment had become, in practice, a medical severity contest.

The costs were not only financial. Workers who experienced discrimination learned that the legal system would first interrogate their bodies before examining their employers’ conduct. The discovery process in a threshold fight is intrusive by nature: medical records, physician depositions, detailed questioning about daily routines and bodily functions. For a plaintiff whose condition was under control, the process offered humiliation without progress, an examination of limitations she worked daily to overcome, conducted for the purpose of proving she was limited enough to deserve protection. The 2008 act’s directive against extensive analysis can be read as a response to this human cost as well as to the doctrinal one. Congress did not only want fewer threshold dismissals. It wanted the litigation to start in a different place.

Employer organizations, for their part, described the pre-2008 regime in terms of predictability and cost. Business groups argued that a broad definition would sweep in minor, temporary, or fully managed conditions, forcing employers to engage in an interactive process and defend accommodation decisions for workers whose impairments had little practical effect on the job. They warned that reopening the definition would increase the volume of charges filed with the Equal Employment Opportunity Commission, raise the cost of litigating the threshold question in reverse, and expose employers to claims from people who could perform every essential function without any change at all. Those concerns were stated in congressional testimony and in public commentary during the legislative debate, and they reflected a genuine worry: that the line between a medical condition and a protected disability would blur beyond usefulness.

Disability-rights advocates told a different story about the same years, and the difference is worth stating with equal care. In their account, the trilogy did not produce predictability. It produced a regime in which the people most likely to experience discrimination were the least likely to be protected, because the conditions employers most readily acted on, epilepsy, diabetes, mental illness, HIV infection, cancer in remission, were precisely the conditions that medication and treatment rendered manageable enough to fail the Court’s test. Advocates pointed to workers fired after disclosing a diagnosis, passed over after requesting a schedule change for treatment, or terminated when a well-controlled condition produced a single visible episode. The employer’s action was driven by the impairment. The court’s dismissal was driven by the mitigation. That gap, between the reason for the adverse action and the reason for the dismissal, was the catch-22 the 2008 act was written to close.

What was the pre-2009 catch-22 that advocates described?

Advocates described workers whose conditions were serious enough to trigger adverse employment decisions yet, once mitigating measures were considered, too well managed to satisfy the courts’ demanding disability test, leaving them unprotected at the exact moment an employer acted on the impairment.

The politics of repair

The legislative path of the ADA Amendments Act is one of the cleaner bipartisan stories of its era, and the details are worth stating precisely because two different House votes are easy to confuse. The House first passed its own bill, H.R. 3195, on June 25, 2008, by a vote of 402 to 17. The Senate then took up its own version, S. 3406, introduced by Senator Tom Harkin, and passed it by unanimous consent on September 11, 2008. Six days later, on September 17, the House set aside its earlier bill and passed the Senate’s S. 3406 by voice vote, sending the identical text to the White House. The September 17 vote was on the Senate bill, not a second vote on the House bill, and the distinction matters for anyone tracing the legislative history: the enacted text is the Senate’s.

Why two House votes

The two-vote sequence puzzles readers who encounter it for the first time, so the mechanics deserve a plain explanation. The House moved first with its own bill, H.R. 3195, which passed overwhelmingly in June. The Senate then developed its own text, S. 3406, through the stakeholder negotiations, and passed it by unanimous consent in September. At that point the House faced a choice: insist on its own bill and go to conference, or accept the Senate’s text and go to the president. It chose acceptance, passing S. 3406 by voice vote and sending it to the White House eight days later. The choice reflected the negotiations’ success. The Senate text already embodied the compromise both chambers’ stakeholders had accepted, so a conference would have added delay without improving the product.

Senator Tom Harkin introduced the Senate bill, S. 3406, which became the enacted text. The unanimous-consent procedure is reserved for measures no senator wishes to block, and its use here signaled that the stakeholder process had cleared the field. Unanimous consent does not mean every senator loved the bill. It means none thought opposition worth the cost, which in a negotiated compromise is the practical definition of consensus. The House’s voice vote six days later completed the picture: a bill that moved through both chambers without a single recorded vote against its final text.

President George W. Bush signed the bill on September 25, 2008, as Public Law 110-325, and the statute set its own effective date of January 1, 2009, in section 8. The signing carried a symmetry that the press noted at the time. The president who signed the amendments was the son of the president who had signed the original Americans with Disabilities Act in 1990. George H. W. Bush had put his name to the 1990 law on July 26 of that year; eighteen years later, George W. Bush put his name to the law that restored the definition his father’s statute had enacted. The disability community, which had spent nearly a decade losing threshold fights, treated the signing as the closing of a long detour. Business groups, which had negotiated the final text through months of talks with advocates, treated it as a workable compromise rather than a defeat, a framing that helps explain the lopsided votes.

A son signs his father’s statute

The ADA Amendments Act of 2008 carries a symmetry that belongs in any telling of its story. President George W. Bush signed the enactment on September 25, 2008, eighteen years after his father, President George H. W. Bush, signed the original Americans with Disabilities Act on July 26, 1990. The Congressional Record later noted the sequence with plain satisfaction: the son’s signature extended the work the father’s signature had begun. The son’s own public papers record him recalling, in 2002, “when my father signed the ADA into law in 1990,” a memory that framed the 2008 signing as continuity rather than departure. The statute is Public Law 110-325, recorded at 122 Stat. 3553, and its eighth section sets the effective date: January 1, 2009.

The eighteen-year gap between the two signatures maps neatly onto the life cycle of the definition at issue. In 1990 Congress wrote a broad definition and said it meant to cover a large population. Across the following decade the Supreme Court read that definition narrowly in a sequence of employment cases, and the lower courts applied those readings with increasing strictness. By 2008 the original coalition that had passed the 1990 statute, disability advocates, business groups, and legislators from both parties, had reconverged around the judgment that the courts had strayed. The son’s signature therefore closed a circle the father’s signature had opened: the same office, the same statute, the same stated purpose, with a judicial detour in between that Congress set out to erase.

The effective date deserves emphasis because it marks the boundary of the new law’s application. Section 8 of the enactment provides that the amendments become effective on January 1, 2009, and the Equal Employment Opportunity Commission’s guidance confirms that the new definition does not apply retroactively. Conduct that occurred before that date is judged under the old definition even if the charge or the lawsuit arrived later. The boundary is clean, and it means the pre-2009 case law retains its relevance for any dispute whose facts arose before the new year. The guide takes up the retroactivity question in detail in its own section, because the boundary governs how researchers and litigants must read every case decided on either side of it.

The bookend and what it signified

The father-and-son symmetry was more than a press-cycle curiosity. It gave the amendments a narrative that bipartisan legislation often lacks: a story about continuity across administrations and across generations. The elder Bush’s signing of the 1990 act had been a landmark of his domestic presidency, enacted with broad support and celebrated as a civil rights milestone. The younger Bush’s signing of the 2008 amendments allowed both parties to claim the disability cause without relitigating the 1990 fight. Republicans could present the amendments as fidelity to a Republican president’s legacy. Democrats could present them as the correction of a judicial narrowing they had long opposed. The narrative served everyone, which is one reason no one disrupted it.

The timing mattered as well. September 2008 was a season of economic crisis, with the financial system in acute distress and Congress consumed by emergency legislation. That a disability definition bill moved by unanimous consent and voice vote in that environment testifies to the completeness of the pre-negotiation. Controversial bills do not glide through a crisis Congress. The amendments glided because the controversy had been resolved before the crisis arrived, in the stakeholder talks that converted opponents into co-authors. The lesson is not that disability policy is immune from partisanship. It is that thorough negotiation can make even a significant statutory change feel inevitable by the time it reaches the floor.

For the disability community, the signing closed a chapter that had begun with bewilderment. Advocates in 1999 had expected the trilogy to be a speed bump; by 2002, after Toyota, they understood it was a wall. The years between produced a generation of lawyers trained to litigate the definition first and the discrimination second, a professional deformation the 2008 act was meant to cure. The community’s account of the era, the catch-22 of the well-treated worker, became the legislative history’s account too, carried into the findings that describe the narrowing. When a social movement’s description of its problem becomes Congress’s stated premise for legislation, the movement has won the framing fight completely. The 2008 act is that victory written into the Code.

The negotiations behind those votes deserve a paragraph of their own, because they explain why the final text looks the way it does. Employer organizations and disability advocates did not simply lobby past each other. They sat down, through a series of facilitated discussions, and bargained over the mechanics. The business side wanted limits that would keep the definition from swallowing every minor ailment, and it got several: the ordinary eyeglasses exception, the transitory-and-minor exception to the regarded-as prong, and the rule denying reasonable accommodation to regarded-as-only plaintiffs. The disability side wanted the four core reversals, and it got all of them. The result was a bill that both business groups and the disability community could support, which is why a unanimous Senate and a voice-vote House were possible. Compromise, in this case, produced clarity rather than muddle, because each side traded on the mechanics rather than the principle.

The facilitated format mattered. Adversarial lobbying produces position papers; facilitated negotiation produces text. The participants worked through the statutory language line by line, testing each provision against the other’s concerns. When business groups worried that disregarding mitigating measures would cover trivial conditions, the negotiators did not answer with rhetoric. They wrote the eyeglasses exception and the transitory-and-minor rule. When advocates worried that the exceptions would swallow the reversals, the negotiators narrowed the exceptions to their precise terms: ordinary eyeglasses only, transitory and minor together, accommodation withheld only from regarded-as-only plaintiffs. The line-by-line method is visible in the final statute’s specificity. General principles would have left the compromises vague and the courts to sort them out. Specific provisions left the courts with instructions, which was the whole point of the exercise.

How did the politics of the amendments unfold?

The House passed H.R. 3195 by 402 to 17 on June 25, 2008, the Senate passed S. 3406 by unanimous consent on September 11, the House then passed the Senate bill by voice vote on September 17, and President George W. Bush signed it on September 25 with an effective date of January 1, 2009.

What each side won at the table

The final text reads like a contract because it was negotiated like one, and the terms of the trade are visible in the statute. The disability community won the four reversals that form the core of the act: mitigating measures disregarded, broad construction of substantially limits, episodic and in-remission coverage, and major bodily functions in the definition. Those were the non-negotiable demands, the provisions without which advocates would not have supported the bill. Everything else in the act is scaffolding around those four.

Employer organizations won the boundaries. The ordinary eyeglasses and contact lenses exception kept routine vision correction outside the protected class, answering the business fear that the restored definition would sweep in tens of millions of nearsighted workers. The transitory-and-minor exception to the regarded-as prong gave employers a defense against claims built on trivial, short-lived conditions. And the rule withholding reasonable accommodation from regarded-as-only plaintiffs ensured that the broadened third prong would function as an anti-stereotyping provision rather than as a new source of accommodation claims. Each of these limits was priced into the negotiations. Advocates accepted them because the core reversals were worth the boundaries, and business groups accepted the core reversals because the boundaries made them administrable.

The negotiation also explains the speed of the final legislative sequence. By the time S. 3406 reached the Senate floor, the hard bargaining was finished, which is why unanimous consent was possible: no senator needed to object because no stakeholder was still fighting. The House’s voice vote six days later reflected the same dynamic. Voice votes are used when the outcome is not in doubt, and the outcome was not in doubt because the coalition held. The lesson for students of legislation is that the lopsided votes were the product of the negotiations, not a substitute for them. The consensus was manufactured before the roll was called, in the months of facilitated discussion that the vote totals merely ratified.

What did employers gain in the negotiated compromise?

Employers gained three boundaries: the ordinary eyeglasses exception, the transitory-and-minor exception to regarded-as coverage, and the rule denying reasonable accommodation to regarded-as-only plaintiffs. Each limit made the restored definition administrable, which is why business groups supported a bill that reopened coverage.

Findings, purposes, and the unusual explicitness

The 2008 act’s section 2 does something Congress rarely does. It does not merely state findings and move on. It builds a record of disagreement with the judiciary, names the decisions, and then states purposes that function as instructions for how the statute must be read. The findings in section 2(a) include, at paragraph (4), the statement that the holdings of the Supreme Court in Sutton and its companion cases and in Toyota Motor Manufacturing v. Williams narrowed the broad scope of protection intended to be afforded by the 1990 law. Paragraph (7) adds related findings about the narrowing. But the findings describe. The rejecting is done in the purposes.

Section 2(b) states Congress’s purposes, and subsections (2) through (4) do the operative work. They reject the requirement enunciated in Sutton and its companion cases that mitigating measures be considered in determining disability. They reject the Court’s reasoning in Sutton with regard to coverage under the third prong, the regarded-as prong, and in doing so reinstate the reasoning of an earlier decision, School Board of Nassau County v. Arline, 480 U.S. 273 (1987), which had taken a broad view of regarded-as coverage. And they reject the standards enunciated in Toyota that the terms “substantially” and “major” need to be interpreted strictly to create a demanding standard for qualifying as disabled. Section 2(b)(5) adds a directive with practical bite for lower courts: the definition of disability should not demand extensive analysis. The message to judges was unmistakable. Stop litigating the threshold so hard.

The attribution matters for anyone reading the statute closely. It is the purposes, not the findings, that carry the express rejections, and drafters and litigators who cite the act should place the “reject” language in section 2(b) rather than section 2(a). The distinction is more than pedantic. Findings describe Congress’s view of the world; purposes state what Congress meant the statute to do. When a court asks what rule replaces Sutton, the answer sits in the purposes and in the amended definitional text of 42 U.S.C. 12102, not in the prefatory findings.

Reading section 2 as a litigator

Section 2 rewards the kind of close reading that practitioners give to operative text, because it is operative text wearing the clothes of preamble. A litigator citing the act for the mitigating-measures rule has three layers to invoke, and each does different work. The purposes in section 2(b)(2) supply the congressional rejection of the Sutton requirement, the answer to any argument that the old rule survives in some interstitial form. The amended definition in section 12102(4)(E) supplies the replacement rule, the positive command to disregard mitigating measures. And the findings in section 2(a)(4) supply the narrative context, Congress’s statement that the holdings narrowed the intended scope, useful for the introduction of a brief but not for the doctrinal point.

The same three-layer structure applies to the Toyota rejection. Section 2(b)(4) rejects the demanding standard. Section 12102(4)(A) and (B) installs broad construction in its place. Section 2(a) describes the narrowing that made the rejection necessary. And for the third prong, section 2(b)(3) rejects Sutton’s reasoning and restores Arline, while section 12102(3)(A) codifies the restored test. The pattern is consistent across all four changes: purposes reject, Code replaces, findings explain. A brief that cites only the findings has the story right but the authority wrong. A brief that cites only the Code has the rule right but misses the interpretive direction the purposes supply. The complete citation uses all three, in that order.

This structure also explains why the act has generated so little threshold litigation since 2009. Override statutes fail when they announce principles and leave courts to work out the details, because courts working out details recreate the old doctrines under new names. The 2008 act does not leave the details to anyone. It names the decisions, rejects their reasoning in the purposes, writes the replacement rules into the Code with illustrative lists, and tells the implementing agency to construe broadly. There is simply less for a court to do, which was the point. Section 2(b)(5)’s directive against extensive analysis is not only an instruction about this definition. It is a confession about the previous decade: Congress watched courts spend nine years analyzing the threshold and decided the analysis itself was the problem.

The override table

The act’s operative provisions translate the purposes into four mechanical changes plus the regarded-as restructuring. The table below sets each rejected holding beside the statutory language that displaces it and the practical effect on who is covered. It is the article’s central artifact, because the 2008 act is best understood not as a set of abstract principles but as a point-by-point reversal. Each row pairs a judicial holding with the Code provision that answers it, so the reader can see the override working at the level of text rather than theme. The holdings are stated from the opinions themselves, not only through the findings’ characterization, and the statutory language is quoted or closely paraphrased from the amended sections. The practical-effect column translates the legal change into coverage terms: who is now protected that was not protected before. Read row by row, the table is the legislative history in miniature. Read as a whole, it shows a Congress that left nothing to inference.

Judicial holding rejected	Statutory language rejecting it	Practical effect on who is covered
Sutton, Murphy, and Albertson’s: mitigating measures, including medication, devices, and the body’s own systems, must be considered in deciding whether an impairment substantially limits a major life activity, 527 U.S. 471, 516, 555 (1999)	42 U.S.C. 12102(4)(E): the disability determination shall be made without regard to the ameliorative effects of mitigating measures, including medication, prosthetics, hearing aids, assistive technology, and learned behavioral or adaptive neurological modifications, with an exception for ordinary eyeglasses and contact lenses	People whose conditions are controlled by treatment, such as epilepsy managed by medication or diabetes managed with insulin, are assessed in their unmitigated state and qualify as disabled, ending the rule that good treatment removed protection
Toyota v. Williams: “substantially” and “major” must be interpreted strictly to create a demanding standard, requiring an impairment that prevents or severely restricts activities of central importance to most people’s daily lives, 534 U.S. 184 (2002)	42 U.S.C. 12102(4)(A) through (B): “substantially limits” shall be construed in favor of broad coverage to the maximum extent permitted, and interpreted consistently with the findings and purposes of the 2008 act; section 2(b)(4) rejects Toyota’s demanding standard, and the EEOC regulations state the standard is not meant to be a demanding one, 29 C.F.R. 1630.2(j)(1)(i)	The threshold inquiry becomes a broad, non-demanding screen rather than a strict evidentiary gauntlet, so fewer plaintiffs lose at the definition stage and courts spend less effort on extensive threshold analysis
The pre-2008 treatment of episodic conditions: an impairment in remission or between episodes could fail the test because it was not limiting at the time of assessment	42 U.S.C. 12102(4)(D): an impairment that is episodic or in remission is a disability if it would substantially limit a major life activity when active	Conditions such as epilepsy, multiple sclerosis, and cancer in remission qualify based on their active state, so a person need not be symptomatic at the moment of the employment decision to be protected
The pre-2008 narrow reading of major life activities, confined to activities like walking, seeing, and working as construed by courts	42 U.S.C. 12102(2)(B): major life activities include the operation of major bodily functions, with a non-exhaustive list covering the immune system, normal cell growth, digestive, bowel, bladder, neurological, brain, respiratory, circulatory, endocrine, and reproductive functions	Impairments that disrupt internal body systems, such as cancer affecting normal cell growth or diabetes affecting the endocrine system, clearly qualify, bringing a wide range of medical conditions within coverage without stretching the statutory text
Sutton’s narrowing of the regarded-as prong: coverage under the third prong required showing the employer perceived a substantially limiting impairment	42 U.S.C. 12102(3)(A): a plaintiff meets the regarded-as requirement by showing an actual or perceived impairment that motivated a prohibited action, whether or not the impairment limits or is perceived to limit a major life activity, restoring the reasoning of School Board of Nassau County v. Arline, 480 U.S. 273 (1987); 42 U.S.C. 12102(3)(B) excepts impairments that are both transitory and minor	A worker fired because an employer acted on a perceived impairment can proceed without proving the impairment was substantially limiting, while employers retain the transitory-and-minor exception and owe no reasonable accommodation to regarded-as-only plaintiffs under 42 U.S.C. 12201(h)

Change one: mitigation set aside

The first mechanical change is the most intuitive and, for practitioners, the most consequential. Section 12102(4)(E) provides that the determination of whether an impairment substantially limits a major life activity shall be made without regard to the ameliorative effects of mitigating measures. The statute then lists examples: medication, medical supplies, equipment and appliances, low-vision devices (excluding ordinary eyeglasses and contact lenses), prosthetics including limbs and devices, hearing aids and cochlear implants, mobility devices, oxygen therapy equipment, assistive technology, reasonable accommodations and auxiliary aids, and learned behavioral or adaptive neurological modifications. The list is illustrative rather than exhaustive, and its breadth shows Congress’s intent to reverse not only Sutton and Murphy but also Albertson’s, whose holding about the body’s own systems is answered by the inclusion of learned behavioral and adaptive neurological modifications.

The exception proves the rule’s deliberateness. Ordinary eyeglasses and contact lenses are carved out: their ameliorative effects are still considered. The carve-out was a negotiated compromise, and its logic is worth stating. Corrective lenses are ubiquitous, cheap, and fully effective for most wearers, and neither side of the negotiations wanted every nearsighted person in America swept into the protected class by a statute aimed at serious disability discrimination. The exception is narrow by design. It covers ordinary eyeglasses and contact lenses, not low-vision devices, and the distinction has generated its own line of interpretive questions. But the core rule is simple to state and to apply: assess the impairment as it would be without the mitigating measure, and decide coverage on that basis.

For employment practice, the change removed an entire category of threshold motion. Before 2009, a defendant could win by showing the plaintiff’s medication worked. After the change, that showing is legally irrelevant to coverage, and the litigation moves to the questions the statute always meant to foreground: can the person perform the essential functions of the job, and would an accommodation be reasonable. The mitigating-measures rule also interacts with the reasonable accommodation duty in a way worth noting. A person whose condition is fully controlled by medication may still need an accommodation, such as a schedule adjustment for treatment, and the 2008 act ensures that the success of the treatment does not disqualify the person from asking.

How the mitigating-measures rule works across cases

The statutory list of mitigating measures repays close reading because its categories map onto the trilogy’s holdings one by one. Medication answers Murphy. Prosthetics, hearing aids, and mobility devices answer the assistive-technology cases that followed Sutton. Assistive technology and auxiliary aids reach the workplace tools that enable performance. And learned behavioral or adaptive neurological modifications answer Albertson’s directly, closing the door the Court had opened to the body’s own systems. Congress did not merely reverse the bottom line of the three decisions. It walked through the categories the decisions had blessed and reversed each one, leaving future courts no room to revive the rule under a new label.

Two interpretive details deserve attention. First, the statute disregards the ameliorative effects of mitigating measures, which means it also disregards their side effects for coverage purposes only in the sense that the side effects are part of the unmitigated picture. A medication that controls seizures but causes drowsiness presents the impairment as it would exist without the drug, and the analysis proceeds from there. Second, the provision says nothing about measures a person could take but has not. The question is the effect of measures actually in use, not the effect of hypothetical treatment. A plaintiff is not required to mitigate in order to qualify, and a defendant cannot argue that the plaintiff would not be disabled if she tried a different regimen. The rule looks at the person as she is, minus the help she actually receives.

The eyeglasses exception, narrow as it is, illustrates the difference between a compromise and a loophole. Ordinary eyeglasses and contact lenses are the only mitigating measures whose effects still count, and the exception reflects a legislative judgment that routine vision correction is too common and too complete to trigger a civil rights statute. Everything else, from low-vision devices to cochlear implants to learned behavioral modifications, is disregarded. The line is deliberately bright. Congress preferred a clear rule with one bargained exception over a flexible standard that courts could stretch back toward Sutton.

Change two: a broad reading of substantially limits

The second change answers Toyota directly. Section 12102(4)(A) provides that the definition of disability shall be construed in favor of broad coverage of individuals, to the maximum extent permitted by the terms of the chapter. Section 12102(4)(B) adds that the definition shall be interpreted consistently with the findings and purposes of the 2008 act, which, as noted, expressly reject the demanding standard. Together, the two subsections convert “substantially limits” from a strict judicial test into a broad statutory screen.

A note on attribution is required here, because a frequently quoted phrase does not come from the statute. The sentence that “substantially limits” is not meant to be a demanding standard appears in the Equal Employment Opportunity Commission’s regulations implementing the act, at 29 C.F.R. 1630.2(j)(1)(i), and in the Commission’s interpretive guidance. The statute itself mandates broad construction and rejects Toyota’s demanding standard through section 2(b)(4); the memorable phrasing is the agency’s. Writers who attribute the phrase to the Code misstate the source. The distinction matters because regulations can be revised by later administrations while the statutory mandate endures, and a careful reader should know which is which.

The practical effect of the broad-construction rule is to demote the threshold inquiry. The statute itself, in section 2(b)(5), directs that the definition should not demand extensive analysis, and the EEOC’s regulations carry that instruction into enforcement. For courts, the message is to stop treating the definition as a demanding standard and to let cases proceed to the merits. For employers, the message is parallel: the question worth analyzing is not whether the worker is disabled enough, but whether the worker can do the job and what change would make that possible.

The two subsections working together

Sections 12102(4)(A) and (4)(B) do different work, and reading them together shows the care of the drafting. Subparagraph (A) states the interpretive direction: construe the definition in favor of broad coverage, to the maximum extent permitted by the chapter’s terms. It is a thumb on the scale, an instruction that close questions about the meaning of substantially limits should be resolved toward coverage. Subparagraph (B) states the interpretive loyalty: interpret the definition consistently with the findings and purposes of the 2008 act. Since the purposes expressly reject Sutton and Toyota, subparagraph (B) functions as a pointer to the rejection, ensuring that no court can construe substantially limits broadly in the abstract while smuggling the old demanding standard back through a different door.

The phrase “to the maximum extent permitted by the terms of this chapter” sets the outer boundary of the instruction. Broad construction is not boundless construction. The definition still requires an impairment, still requires a substantial limitation, and still requires a major life activity or bodily function. A plaintiff must still show something real. What the instruction forbids is the Toyota approach of reading each of those requirements strictly so that their combined effect is to demand an extraordinary showing. The 2008 act keeps the architecture of the definition and changes the attitude with which it is read. That is why the provision is both modest and transformative: modest in what it adds to the Code, transformative in what it tells judges to stop doing.

How should a reader apply the amended definition step by step?

First identify the impairment, then disregard any mitigating measures except ordinary eyeglasses, then ask whether it would substantially limit a major life activity or bodily function when active, construing substantially limits broadly. If the answer is yes, move immediately to qualification and reasonable accommodation.

Change three: the episodic and in-remission rule

The third change is the shortest in the Code and among the most humane in effect. Section 12102(4)(D) provides that an impairment that is episodic or in remission is a disability if it would substantially limit a major life activity when active. The rule answers a problem the trilogy had sharpened: conditions that flare and recede, such as epilepsy, multiple sclerosis, bipolar disorder, and many cancers, could fail the pre-2008 test during a quiescent period, even though the employer’s adverse action was often triggered by an episode or by knowledge of the diagnosis.

The provision works by shifting the temporal frame. The question is not whether the impairment limits the person today, at the moment of the challenged decision, but whether it would substantially limit a major life activity when active. A worker with epilepsy whose seizures are infrequent, a cancer survivor in remission, and an employee with a mental health condition managed between episodes are all assessed on the active state. The rule also dovetails with the mitigating-measures provision: medication that keeps a condition in remission is disregarded, and the remission itself does not defeat coverage. Together, the two provisions ensure that the natural history of a chronic condition, flare, treatment, remission, cannot be used to define the person out of the protected class.

Why the temporal frame matters

The episodic-or-in-remission rule solves a problem that is easy to miss until it is stated plainly. Discrimination often happens precisely when a condition is not active. An employer learns of a past cancer diagnosis during a background check, or hears about an employee’s epilepsy from a coworker, and acts on the knowledge while the worker is perfectly healthy. Under the pre-2008 framework, the employer’s timing created a defense: the plaintiff was not substantially limited at the moment of the adverse action, so the definition failed. The statute thus protected people least at the moments they most needed protection, when an employer acted on history or diagnosis rather than on present incapacity.

The 2008 rule breaks that connection. By asking what the impairment would do when active, it aligns the legal test with the employer’s actual motivation. If the employer acted because of the epilepsy, the relevant question is what epilepsy does when active, not what the worker’s Tuesday looked like. The provision also protects the record-of prong’s neighbors: a cancer survivor in remission may proceed under the first prong directly, without needing to frame the claim as discrimination based on a record of past impairment. The drafters thus gave episodic conditions two doors into coverage, the active-state rule and the record-of prong, reflecting a judgment that the pre-2008 law had left these plaintiffs with none.

Change four: major bodily functions

The fourth change expands the second half of the first prong. Section 12102(2) defines major life activities, and subparagraph (B) adds that the term includes the operation of a major bodily function. The statute then gives a non-exhaustive list: functions of the immune system, normal cell growth, digestive, bowel, bladder, neurological, brain, respiratory, circulatory, endocrine, and reproductive functions. The list is illustrative, signaled by the word “including,” so courts may recognize additional bodily functions as the science develops.

The provision’s significance is easiest to see through examples. Before 2008, a plaintiff with cancer had to show that the disease substantially limited a traditional major life activity such as caring for oneself or working, a showing that could fail for a person who continued working through treatment. After the change, cancer’s effect on normal cell growth supplies the major life activity directly. Diabetes’s effect on the endocrine system, HIV’s effect on the immune system, and epilepsy’s effect on neurological and brain function each map onto the statutory list without strain. The provision brought a wide range of serious medical conditions clearly within coverage, and it did so through the statutory text rather than through judicial inference, which was precisely Congress’s method throughout the act.

The non-exhaustive character of the list deserves emphasis because it answers a recurring misreading. Some early commentary treated the list as closed, as though only the named functions counted. The statute says “including,” and the EEOC’s guidance treats the list as open. The open texture is deliberate. Congress in 2008 could not have enumerated every bodily function that future medicine would understand, and it wrote the provision to accommodate that ignorance. The functions named are examples of the principle, not its boundary.

Why bodily functions instead of named conditions

Congress could have restored coverage by listing conditions: epilepsy, diabetes, cancer, HIV, multiple sclerosis, bipolar disorder. It chose not to, and the choice reflects a durable legislative wisdom. A list of conditions freezes the statute at the medicine of its enactment year. New diagnoses emerge, existing diagnoses are reclassified, and a condition-based list would require constant amendment to keep pace. By defining the protected class functionally, through the body systems an impairment disrupts, Congress wrote a definition that absorbs medical progress without further legislation. The question is never whether the condition appears on a list. It is whether the condition substantially limits a listed or unlisted bodily function, assessed broadly and without regard to mitigation.

The functional approach also avoids a hierarchy of diagnoses that a condition list would create. Had Congress named some conditions and omitted others, courts would have read the omissions as deliberate, and advocates for unnamed conditions would have faced an uphill interpretive fight. The bodily-functions provision treats all serious disruptions of body systems alike, whether the disruption comes from a famous disease or an obscure one. That evenhandedness is part of what makes the provision a restoration rather than a favor to particular constituencies. It does not pick winners among diagnoses. It describes how bodies work and protects people whose bodies work differently.

The provision’s relationship to the other three changes completes the picture. The mitigating-measures rule says to ignore treatment. The episodic rule says to assess the active state. The bodily-functions rule says where to look for the limitation. Together, they form a coherent method: take the impairment as it would be without help, at its most limiting, and ask whether it substantially disrupts how the body functions, reading substantially in favor of coverage. A reader who internalizes that method can apply the amended definition to any condition, including ones Congress never discussed, which is exactly what the drafters intended.

Why does the statute list bodily functions instead of naming conditions?

A condition list would freeze coverage at the medicine of 2008 and force courts to treat omissions as deliberate. The functional approach protects any serious disruption of body systems, absorbs future medical understanding without new legislation, and avoids ranking diagnoses against one another.

The regarded-as trade

The restructuring of the third prong is the act’s most elegant piece of drafting, because it pairs an expansion with a limitation in a single bargain. Section 12102(3)(A) provides that an individual meets the regarded-as requirement if she establishes that she was subjected to an action prohibited by the chapter because of an actual or perceived physical or mental impairment, whether or not the impairment limits or is perceived to limit a major life activity. The plaintiff no longer needs to show that the employer regarded her as substantially limited. She needs to show only that the employer acted because of an actual or perceived impairment. The provision restores the reasoning of School Board of Nassau County v. Arline, 480 U.S. 273 (1987), the pre-Sutton decision that had taken a broad view of the prong, and the purposes in section 2(b)(3) say so expressly.

The trade is in the two provisions that follow. Section 12102(3)(B) creates an exception for impairments that are transitory and minor, with transitory defined as an actual or expected duration of six months or less. An employer who acts on the basis of a truly minor and short-lived condition, a sprained ankle expected to heal in weeks, for example, does not trigger regarded-as coverage. And section 12201(h) provides that a covered entity need not provide a reasonable accommodation or reasonable modification to an individual who meets the definition of disability solely under the regarded-as prong. The logic is coherent. The prong targets stigma and stereotype, the employer’s reaction to a perceived impairment, and the remedy for that wrong is to stop the discriminatory action, not to restructure the job around an impairment the person may not even have.

For practitioners, the regarded-as prong after 2008 is a cleaner tool than it was before. It captures the supervisor who fires a worker after learning of a past cancer diagnosis, the hiring manager who rejects an applicant with a visible but non-limiting condition, and the employer who acts on a mistaken belief about a medical test. It does not capture every slight, because the transitory-and-minor exception screens out the trivial, and it does not create an accommodation entitlement, because section 12201(h) withholds one. The bargain reflects the negotiations that produced the act: advocates got a usable third prong, and employers got boundaries around it.

How did the regarded-as prong change?

The 2008 act lets a plaintiff satisfy the prong by showing an employer acted because of an actual or perceived impairment, without proving the impairment substantially limits anything, while adding a transitory-and-minor exception and withholding reasonable accommodation from regarded-as-only plaintiffs.

The third prong, before and after

The contrast between the old and new regarded-as tests illustrates the statute’s method more sharply than any other provision. Under Sutton, the prong asked what the employer believed about the severity of the impairment: the plaintiff had to show that the employer regarded the impairment as substantially limiting a major life activity. The test made the employer’s subjective theory of medical severity the decisive fact, and employers rarely supplied one. An employer who rejected the Sutton sisters for failing a vision standard, or who dismissed Murphy for failing a federal certification, could truthfully say it had no view at all about whether the impairment substantially limited a major life activity. It had a view about the job requirement, and that was not enough.

Under section 12102(3)(A), the prong asks what the employer did because of the impairment: the plaintiff must show an actual or perceived impairment and a prohibited action taken because of it. The employer’s theory of severity drops out entirely. The same vision standard that defeated the sisters’ claim under Sutton now supports a claim under the new test, because the action was taken because of the perceived impairment whether or not anyone believed it substantially limited anything. The provision restores the Arline approach the purposes cite, in which the Court had treated the employer’s perception and the consequent exclusion as the operative facts.

The two qualifications keep the rebuilt prong within bounds. The transitory-and-minor exception in section 12102(3)(B) preserves a de minimis boundary: impairments expected to last six months or less that are also minor do not support coverage. The no-accommodation rule in section 12201(h) preserves the logic of the accommodation duty: a plaintiff covered only because the employer acted on a perception cannot demand changes to address a limitation that does not actually exist. Together the qualifications show the statute’s characteristic balance. Congress widened the prong dramatically and then fenced it with the two limits the legislative bargain required.

The prong’s deeper logic

The regarded-as prong has always been the odd one out in the definition, because it protects people who may not be functionally limited at all. Its target is the employer’s mind, not the worker’s body. The wrong it addresses is decision-making driven by fear, stereotype, or misinformation about a condition: the manager who assumes a worker with a history of depression cannot handle pressure, the interviewer who sees a prosthetic limb and imagines incapacity, the supervisor who learns of an HIV diagnosis and invents a safety concern. In each case, the harm flows from the perception, and the statute’s answer is to forbid acting on it.

Sutton’s narrowing of the prong had broken that logic by reintroducing a functional test through the back door. If the plaintiff had to show the employer perceived a substantially limiting impairment, then the prong protected only people whose perceived conditions were serious enough to satisfy the demanding standard, which was precisely the showing most stigma cases could not make. A manager who fires a worker over an unfounded fear does not perceive a substantially limiting impairment in the Toyota sense. She perceives a threat, a strangeness, a risk she cannot articulate. The 2008 act restores the prong to its proper function by asking only whether the employer acted because of an actual or perceived impairment. The functional inquiry drops out because it was never the point.

The two boundaries keep the restored prong honest. The transitory-and-minor exception ensures that the prong does not become a general civility code for every passing ailment an employer notices. The six-month line gives employers a predictable shelter for genuinely trivial conditions while leaving the serious cases covered. And the withholding of reasonable accommodation from regarded-as-only plaintiffs preserves the prong’s identity as an anti-discrimination rule rather than an accommodation rule. A worker who is impaired enough to need changes at work proceeds under the first prong, where the accommodation duty lives. A worker whose only claim is that the employer acted on a perception gets the discrimination claim without the accommodation claim. The structure is coherent because each prong does the work it was designed to do.

The agency’s implementing rules

Statutes do not administer themselves, and the 2008 act left important details to the Equal Employment Opportunity Commission, which enforces the employment title. The Commission’s regulations at 29 C.F.R. part 1630 translate the four mechanical changes into operational guidance for investigators, employers, and courts. The most quoted provision, section 1630.2(j)(1)(i), states the rule for construing substantially limits: the term shall be construed broadly in favor of expansive coverage, and it is not meant to be a demanding standard. As noted earlier, that phrasing is the agency’s, and the attribution matters. But the regulation’s substance tracks the statute faithfully, and courts give the Commission’s interpretation respectful consideration.

The regulations also carry forward the statute’s instruction against extensive threshold analysis, directing that the coverage question be treated as a broad screen rather than an evidentiary gauntlet. The practical consequence for employers and investigators is a simpler inquiry: apply the four statutory rules, construe substantially limits broadly, and move on to qualification and accommodation. A human resources department working under the new framework spends its analytical effort on the interactive process rather than on disputing whether a worker with a serious medical condition counts as disabled.

The Commission’s questions-and-answers on the final rule address the temporal question as well, taking the position that the amendments do not apply retroactively. Conduct before January 1, 2009 is judged under the old definition; conduct on or after that date is judged under the new one. The Q&A format matters because it shows the agency thinking like a practitioner: the abstract question of retroactivity becomes a concrete instruction about which definition to apply to a December termination charged in February. The agency’s answer is the bright line described earlier, and it has governed charge processing since the effective date.

The parallel statute brought along

The 2008 act did not confine itself to the Americans with Disabilities Act. Section 7 of the act amended the definitional provisions of the Rehabilitation Act of 1973, the earlier funding-condition statute that bars disability discrimination by federal agencies, federal contractors, and recipients of federal financial assistance. The conforming amendments struck the Rehabilitation Act’s freestanding language and tied its definition of disability to the definition in section 3 of the 1990 act, at 29 U.S.C. 705(9)(B) and 29 U.S.C. 705(20)(B). The effect was to carry the four mechanical changes and the regarded-as restructuring into the Rehabilitation Act’s coverage as well.

The conformity matters for two reasons. First, it prevents a split regime in which a federal employee or a student at a federally funded university faces a different definition of disability than a private-sector worker. Before the conforming amendments, the Rehabilitation Act’s definition had developed its own case law, and the Supreme Court’s narrowing decisions had influenced its interpretation too. Tying the two definitions together ensures that the override reaches every forum where the federal disability definition operates. Second, the conformity illustrates a drafting habit worth noticing: when Congress repairs a definition, it repairs it everywhere the definition appears. The 1973 funding-condition statute that the 2008 act amended has its own history, and the conforming amendments are the bridge between the two laws. Readers working through the before-and-after holdings may find a legislation study notebook useful for tracking which provision answers which decision.

The conformity in practice

The two paragraphs amended, 29 U.S.C. 705(9)(B) and 705(20)(B), cover the Rehabilitation Act’s core definitional provisions, and amending both ensured that no corner of the older statute retained the pre-2008 understanding. The choice to tie the definition to section 3 of the 1990 act, rather than rewriting the Rehabilitation Act’s language independently, was a decision for perpetual conformity. If Congress amends the ADA definition again, the Rehabilitation Act follows automatically. The drafters preferred a living link over a snapshot, and the preference reveals their diagnosis of the problem: definitions drift when they are stated in two places, and the way to prevent drift is to state them once.

For the people the Rehabilitation Act protects, the practical effect mirrors the ADA’s. A federal employee with a mitigated condition is assessed without regard to the mitigation. A student whose episodic condition is in remission is assessed on the active state. A federal contractor’s applicant with a condition affecting a major bodily function proceeds under the expanded definition. The forums differ, administrative complaints, federal-sector EEO processes, and lawsuits with their own procedural rules, but the threshold question now receives the same answer everywhere. The 2008 act thus did more than fix one statute’s definition. It unified the federal definition of disability across the two great disability civil rights laws, so that a worker’s coverage does not depend on which statute her employer falls under.

The statute’s architecture: where each change lives

A reader who wants to find the 2008 act in the Code should understand how the amendments sit within 42 U.S.C. 12102, because the section’s architecture explains how the pieces interact. The section opens with the definition itself: disability means a physical or mental impairment that substantially limits one or more major life activities, a record of such an impairment, or being regarded as having such an impairment. That three-part sentence is the 1990 text, unchanged. Everything the 2008 act added hangs below it as rules of construction, and the placement is deliberate. Congress did not rewrite the definition’s words. It rewrote the instructions for reading them.

Paragraph (2) defines major life activities, and it is here that the bodily-functions expansion lives, in subparagraph (B). Paragraph (3) defines the regarded-as prong, with subparagraph (A) stating the new test, actual or perceived impairment behind a prohibited action, and subparagraph (B) stating the transitory-and-minor exception. Paragraph (4) carries the rules of construction: subparagraph (A) mandates broad construction, subparagraph (B) ties interpretation to the findings and purposes, subparagraph (D) states the episodic-or-in-remission rule, and subparagraph (E) states the mitigating-measures rule with its eyeglasses exception. The section thus reads as a definition followed by a user’s manual, and the manual is longer than the definition because the manual is where the override happened.

The architecture has a practical consequence for research. A lawyer who reads only the opening definition will miss the entire 2008 act, because the opening sentence is the 1990 text. The amendments live in the paragraphs below, in the rules that tell courts how to apply the sentence above. This is why the act is sometimes misdescribed as a minor technical correction: its operative provisions are labeled as construction rules rather than as new rights. But construction rules are where coverage lives or dies, as the Sutton-to-Toyota years proved. Congress understood that the definition’s meaning had always resided in the instructions for reading it, and it rewrote the instructions rather than the words.

The defenses that survived intact

Because the amendments left every defense untouched, the post-2009 practice of disability law is largely the pre-2009 practice minus the threshold fight, and the surviving defenses deserve a clear statement. An employer may still show that the worker is not qualified, meaning unable to perform the essential functions of the position with or without reasonable accommodation. The qualification inquiry is where disputes about job requirements now concentrate, and it is fact-intensive in the way the threshold inquiry was not: it asks what the job actually requires, not what the worker’s body can do in the abstract.

An employer may still show undue hardship, meaning that a requested accommodation would impose significant difficulty or expense relative to the employer’s resources. The undue hardship defense is the statute’s safety valve for genuine burden, and it survived the amendments without alteration. An employer may still invoke direct threat, the defense that a worker poses a significant risk to the health or safety of others that cannot be eliminated by reasonable accommodation. None of these defenses was narrowed, broadened, or restated by the 2008 act.

The survival of the defenses is what makes the restoration framing accurate rather than merely rhetorical. Had Congress meant to expand employer obligations, the defenses would have been the place to do it, by raising the undue hardship bar or narrowing direct threat. Congress did neither. It left the entire apparatus of employer protection in place and changed only the doorway. A defense lawyer advising a client after 2009 thus gives advice that would have been familiar in 1995: document the essential functions, engage in the interactive process in good faith, assess hardship honestly, and do not act on stereotypes. The only advice that changed is the threshold advice, and it changed completely: stop litigating whether the worker is disabled, and start litigating whether the decision was justified.

Three misreadings the text forecloses

The 2008 act is short enough to read in full, yet three misreadings recur in commentary, and the text forecloses each of them plainly. They are worth naming because they track the brief’s recurring errors, the mistakes even careful readers make.

The first misreading is that the act changed what employers must do. It did not. The reasonable accommodation duty, the qualification standard, and every defense remain exactly as the 1990 statute set them. A reader who comes away believing that the amendments created new obligations has confused the doorway with the house. More people can now enter, but nothing inside has moved. The practical consequence is that pre-2009 case law on accommodation, essential functions, and undue hardship remains good law. Only the threshold cases, the ones about who counts, belong to a closed era.

The second misreading is that regarded-as plaintiffs gained accommodation rights along with their broadened coverage. They did not. Section 12201(h) is explicit: no reasonable accommodation or modification is owed to someone covered solely under the third prong. The confusion is understandable, because the prong’s expansion looks like a general enlargement of rights, but the statute pairs the expansion with the limitation in a single design. A plaintiff who wants an accommodation must satisfy the first prong. The third prong offers a discrimination claim and nothing more.

The third misreading is that the broader definition reaches backward. It does not. The act took effect January 1, 2009, and the EEOC’s position is that it does not apply retroactively. A worker terminated in December 2008 cannot invoke the new definition, even if the charge is filed in 2009. The line rewards attention because continuing violations and delayed charges can straddle it, but the principle is the agency’s bright line: old conduct, old definition; new conduct, new definition. Each of these misreadings has the same root, the assumption that a bigger protected class means bigger duties. The statute separates the two questions, and careful readers should too.

What the revisions left untouched

The most misunderstood feature of the 2008 act is what it did not do. The act changed the definition of disability and nothing else of substance. It did not alter the reasonable accommodation duty. It did not change the requirement that a plaintiff be qualified, meaning able to perform the essential functions of the job with or without accommodation. It did not modify any defense, including undue hardship, direct threat, or business necessity. It did not change the remedies available under any title. An employer who could have defeated a claim in 2007 by showing the worker was not qualified, or that the requested accommodation imposed an undue hardship, can defeat the same claim after 2009 on the same grounds. The substantive duty is identical. Only the doorway is wider.

This is the point the brief’s complication names directly, and it deserves restatement because the public shorthand gets it wrong. Headlines described the act as expanding disability rights. The accurate description is that it restored the coverage Congress said it had enacted in 1990. The rights themselves, the duty to accommodate, the prohibition on discrimination, the defenses, were untouched. What changed was who gets through the door to assert them. An expansion of obligation would have required employers to do new things. A restoration of coverage requires them to answer claims from people the statute always meant to protect. The difference is not rhetorical. It determines which arguments remain available in litigation, and the answer is all of them except the threshold attack on the definition.

The restoration framing also answers the employer concern about broadened coverage on its own terms. If the 1990 Congress meant to protect people with epilepsy, diabetes, cancer, and HIV, then the 2008 act does not newly burden employers with those workers’ claims. It stops courts from relieving employers of claims the statute always contemplated. The burden employers feel is real, but it is the burden the original bargain imposed, not a new one. Disability advocates make the mirror argument: the pre-2009 years were the anomaly, a period when judicial narrowing suspended the statute’s intended reach, and the amendments merely ended the suspension. Both accounts converge on the same description. The law went back to what it said.

The restoration framing also did political work that an expansion framing could not have done. An expansion invites opponents to ask what is being expanded and at whose expense, and it gives every affected industry a reason to fight. A restoration invites a narrower question: did the courts get the original meaning wrong. The findings answered yes, with citations, and the purposes supplied the correction. Business groups could support a restoration because it came paired with the boundaries they had negotiated; they were not being asked to endorse an open-ended enlargement of duties but to accept a return to the bargain their predecessors had struck in 1990. Disability advocates could support it because the return was real, measured in the four reversals that reopened the doorway. The framing was not spin. It was the accurate description, and its accuracy is what made the coalition possible.

There is a final reason the distinction matters, and it concerns the future. An expansion of obligation invites the question of how far the expansion goes and when it stops. A restoration of coverage invites a different question: whether the restored definition matches what Congress enacted. The first question has no principled stopping point. The second has an answer in the legislative record, the findings and purposes that name the rejected holdings and describe the intended scope. By framing the act as restoration, Congress gave future courts a fixed point to interpret against, rather than an open-ended mandate to keep expanding. The override cycle thus ends where it began, with the text Congress wrote, and the text is now clear enough to hold.

The act’s temporal scope reinforces the point. Section 8 made the act effective January 1, 2009, and the Equal Employment Opportunity Commission’s guidance takes the position that the amendments do not apply retroactively. Conduct that occurred before the effective date is judged under the pre-2008 definition, even if the charge is filed later. A termination in December 2008 is analyzed under Sutton and Toyota; a termination in January 2009 is analyzed under the new text. The line is bright, and it has produced its own body of litigation about which side of the line particular continuing violations fall on. But the principle is settled in the agency’s view: the restoration runs forward, not backward.

What did the amendments leave untouched?

The 2008 act left the accommodation duty, the qualification requirement, every defense, and all remedies exactly as the 1990 statute had originally set them, changing only who satisfies the definition of disability and therefore who may assert the unchanged rights.

The transition across the effective date

The January 1, 2009 effective date created a boundary that every practitioner had to learn to navigate, because cases do not always respect calendar lines. The EEOC’s position is straightforward: the amendments do not apply retroactively, so conduct before the effective date is judged under the old definition and conduct on or after it under the new one. The straightforward statement covers the straightforward cases. A termination in November 2008 is analyzed under Sutton and Toyota. A failure to promote in March 2009 is analyzed under the amended text. The difficulty lies in the cases that straddle the line.

Continuing violations present the hardest straddle problem. A worker subjected to an ongoing pattern of adverse actions beginning in 2008 and continuing into 2009 cannot invoke the new definition for the 2008 acts, even if the charge is filed after the effective date. Each act is judged under the definition in force when it occurred. The principle is the same one the EEOC stated in its questions-and-answers: the definition that governs is the one that governed at the time of the conduct. Practitioners learned to plead and prove accordingly, separating pre-2009 acts from post-2009 acts and recognizing that the same course of conduct could produce different coverage answers on either side of the line.

The transition also affected cases already in the pipeline on the effective date. A lawsuit filed in 2008 but tried in 2009 concerned pre-2009 conduct, so the old definition applied despite the new trial date. A charge filed with the EEOC in late 2008 and investigated in 2009 was processed under the definition in force when the alleged discrimination occurred. The agency’s guidance made the point with a concrete example: an employer that failed to hire, terminated, or denied a reasonable accommodation in December 2008 would have that decision judged under the original definition even if the worker did not file until after January 1, 2009. The example is the EEOC’s, and it illustrates the general rule that filing dates do not move conduct across the boundary. Only the conduct’s date matters.

Over time, the transition questions faded as the pre-2009 conduct aged out of litigation. But the non-retroactivity principle remains part of the act’s identity, because it confirms the restoration framing. A statute that expanded obligations might have been given backward reach to correct past wrongs. A statute that restored a definition applies forward, governing future conduct under the corrected understanding. The bright line is thus of a piece with everything else the act does: it fixes the rule for the future and leaves the past to the law as it was.

The amended definition at work: an illustration

Abstract rules become clear through application, so consider a hypothetical worker and trace her through the definition before and after the 2008 act. Suppose a warehouse employee has epilepsy. Medication controls her seizures almost completely; she has had one breakthrough seizure in three years. Her supervisor learns of the condition, decides she is a safety risk, and reassigns her to a lower-paying role with no safety rationale beyond the diagnosis. She sues.

Under the pre-2008 framework, her case likely died at the threshold. The court would consider her impairment in its mitigated state, well-controlled epilepsy, and ask whether it substantially limited a major life activity under Toyota’s demanding standard. A well-managed condition that produced one seizure in three years would struggle to satisfy a test requiring prevention or severe restriction of activities central to most people’s daily lives. The regarded-as prong offered little help after Sutton’s narrowing, because the employer could argue it did not perceive a substantially limiting impairment. The supervisor’s stereotyped safety judgment, the actual wrong, would go unexamined because the plaintiff was not disabled enough to complain about it.

Under the amended definition, the analysis runs differently at every step. The mitigating-measures rule directs the court to disregard the medication and assess the epilepsy unmitigated. The episodic rule directs assessment on the active state, when seizures substantially limit neurological and brain function, both now listed as major bodily functions. Substantially limits is construed broadly, not as a demanding standard, and the definition should not demand extensive analysis. She is disabled under the first prong. Alternatively, the regarded-as prong covers her directly: the supervisor acted because of an actual impairment, and no showing about substantial limitation is required. The threshold question that consumed the pre-2009 case is answered in a paragraph, and the litigation moves to the merits: was the reassignment discriminatory, was she qualified for the warehouse role, and would any accommodation have been reasonable. The illustration is hypothetical, but the doctrinal path it traces is the one the statute now requires.

The definition across the five titles

The 1990 act’s definition of disability operates across all five titles, and the 2008 amendments travel with it everywhere it goes. The employment title is where the threshold fight was fiercest, because employment cases turn on individualized assessments of workers and jobs, and because the trilogy and Toyota were employment cases. But the restored definition also governs the public services title, the public accommodations title, the telecommunications title, and the miscellaneous provisions. A person with a mitigated condition who challenges a government program’s eligibility rule, or who seeks an auxiliary aid from a place of public accommodation, invokes the same amended definition as the employment plaintiff.

The cross-title reach matters because the pre-2008 narrowing had distorted non-employment cases too. A student with a controlled condition seeking accommodations from a university, or a citizen with an episodic impairment challenging a government policy, faced the same threshold gauntlet as the employment plaintiff. The 2008 act’s restoration is not confined to the workplace. It reopens the doorway in every forum the statute touches, which is why the conforming amendments to the Rehabilitation Act were a natural complement: the federal definition now speaks with one voice across the employment, government services, public accommodations, and federal funding contexts.

The unity has a practical consequence for counseling. A lawyer advising a client with a disability no longer needs to ask which title’s definition applies, because the answer is the same everywhere. The variations that remain concern duties, defenses, and remedies, which differ by title, not the threshold question of who is covered. That simplification is one of the quiet achievements of the 2008 act. Before it, the definition’s meaning could vary with the forum’s case law. After it, the meaning is stated once, in section 12102, and applied everywhere. The statute’s center of gravity shifted from the courts back to the Code.

That shift is worth pausing over, because it reverses the usual direction of legal development. Normally, a statute’s meaning migrates over time from the enacted text toward the accumulated case law, as courts fill gaps and later readers cite the decisions rather than the Code. The 2008 act ran the migration in reverse. It took nine years of accumulated case law, identified the holdings as errors, and replaced them with text, moving the center of authority from the reporters back to the statute book. A lawyer researching the definition starts with section 12102 and finds the answers there, consulting the pre-2009 cases only for historical context. The override thus did more than correct four holdings. It restored the primacy of the enacted text as the source of the law’s meaning, which is where a legislature that disagrees with its courts has the final word.

The battleground moved

The namable claim of this article is also its most practical lesson. Before 2009, employers won most disability cases by arguing the plaintiff was not disabled, and the argument usually succeeded because the trilogy and Toyota had made the definition so hard to satisfy. After the effective date, that argument largely disappeared from motion practice, not because employers stopped raising it but because it stopped working. The contest relocated to the questions the statute had always contained: is the person qualified, and was the requested accommodation reasonable. The litigation moved without any change to the substantive duty, a complete relocation of the battlefield.

The relocation changed the economics and the strategy of disability cases on both sides. Plaintiffs’ lawyers no longer need to front-load medical experts to survive summary judgment on the definition, which lowers the cost of bringing meritorious cases and shifts resources to proving qualification and the reasonableness of the accommodation. Defense counsel, deprived of the threshold kill shot, must engage with the interactive process, document undue hardship, and litigate the essential functions of the job, which are fact-intensive inquiries that resist early disposition. The character of disputes changed even where their number did not: fewer cases about whether the plaintiff counts, more cases about what the employer should have done.

The change also altered the advice lawyers give before anyone files. An employment lawyer evaluating a potential case in the Sutton era began with a medical workup, because the definition was the first and highest hurdle. The same lawyer after 2009 begins with the job: what are its essential functions, what change is the worker requesting, and can the employer show the change is unreasonable or an undue hardship. The intake conversation moved from the doctor’s office to the workplace, which is where Congress meant it to happen all along. For workers, the practical difference is that the conversation with counsel starts with what happened at work rather than with a debate about whether a managed condition is serious enough to mention.

Employers adapted by professionalizing the interactive process, the structured dialogue through which accommodation requests are evaluated. When the threshold defense was available, some employers treated the process as optional, reasoning that a worker who was not disabled had no right to it. After the amendments, the process became the main event, and human resources departments invested in training, documentation, and consistent procedures. That investment is the unglamorous but real consequence of the battleground moving: the fight relocated from the courtroom’s threshold to the conference room, and the conference room had to get better at its job.

Both sides’ accounts of the relocation deserve the same careful attribution given earlier. Employer advocates acknowledge that the threshold fight is largely over and argue that the new fights, over the reasonableness of accommodations and the definition of essential functions, are where the real costs now sit. They contend that the broad definition forces human resources departments to process accommodation requests from workers whose impairments have minimal job impact, and that the interactive process has become a compliance burden detached from genuine need. Disability advocates reply that the new fights are the fights Congress intended in 1990, that arguing about accommodations is arguing about the actual problem rather than about vocabulary, and that the pre-2009 regime wasted years of litigation on a question that served neither workers nor employers. The two accounts disagree about costs and benefits, but they agree on the description: the battleground moved, and it has not moved back.

For workers deciding whether to bring a claim, the relocation changed the calculus in a way that statistics cannot capture but experience confirms. Under the old regime, the threshold fight meant that a worker with a strong discrimination claim but a well-managed condition faced long odds, high expert costs, and intrusive medical discovery before ever reaching the employer’s conduct. Many such workers never filed, and many lawyers declined their cases at intake. Under the new regime, the intake question is whether the employer’s action was discriminatory and whether an accommodation was reasonable, questions that can be evaluated from the workplace facts without a medical gauntlet. The change does not guarantee any worker a victory. It guarantees that the fight, win or lose, will be about what happened at work rather than about whether the worker’s body qualifies for the statute’s attention. That is the relocation’s human meaning, and it is the reason the disability community treats the 2008 act as the most important disability legislation since the 1990 statute itself.

Questions the amendments left open

No statute answers everything, and the 2008 act’s very explicitness throws its remaining ambiguities into relief. The open questions are narrower than the ones the act closed, but they are real, and a candid guide should name them.

The first concerns the reach of the bodily-functions list. The statute says “including,” which signals that unlisted functions count, but it gives no test for recognizing one. Future litigation will ask whether a given physiological process qualifies as a major bodily function, and courts will answer with the broad-construction mandate as their guide. The drafters accepted this residual uncertainty as the price of a future-proof definition. A closed list would have been precise and obsolete within a decade. An open list is durable and requires judgment at the margins.

The second concerns the transitory-and-minor exception. Six months is a bright line for transitory, but minor has no statutory definition, and the two requirements must both be met. Employers and courts will work out, case by case, which short-lived conditions count as minor and which do not. The exception’s narrowness is intentional: it was priced as a small concession in the negotiations, and reading it broadly would upset the bargain. But its boundaries will be litigated, because boundaries always are.

The third concerns the interaction between the prongs. A plaintiff who satisfies the first prong and the third prong simultaneously, the worker with an actual substantially limiting impairment whose employer also acted on stereotype, proceeds with full rights, including accommodation. A plaintiff covered only under the third prong does not. The line between the two will matter in close cases, particularly where the impairment’s functional effects are disputed. These are the ordinary growing pains of a new statutory regime, not defects in the drafting. The 2008 act closed the questions that had paralyzed the field for nine years. The questions it left are the kind that any functioning legal regime generates and resolves over time.

A fourth open area concerns the implementing agency’s ongoing role. The 2008 act directs the Equal Employment Opportunity Commission, along with the other enforcing agencies, to issue regulations carrying out the amendments, and the content of those regulations will shape how the broad-construction mandate applies in new contexts. Regulations can clarify, but they can also become the next site of contest, as each administration revisits the prior one’s guidance. The statute’s own text is the backstop: whatever the regulations say in a given year, the Code mandates broad construction, rejects the demanding standard, and lists the four mechanical rules. The agency layer adds detail, but the statute does not depend on it. That independence is another deliberate feature of the drafting. Congress wrote the override into the Code precisely so that it would not live or die with any single administration’s regulatory choices.

Restoration, not expansion, and the override cycle

The 2008 act belongs to a pattern in American civil rights law that this series has traced across clusters: Congress writes a broad statute, the Supreme Court narrows it, and Congress overrides the narrowing with amendments that restore the original scope. The Civil Rights Act of 1991 is the closest parallel, a statute that reversed a string of late-1980s employment discrimination decisions in the same way the 2008 act reversed the disability-definition decisions. The 1991 amendments that overrode a string of decisions in the employment discrimination cluster show the same legislative rhythm: judicial narrowing, congressional findings of disagreement, and operative text that displaces the holdings point by point. The labor cluster has its own examples. The override cycle is not an anomaly. It is how the dialogue between the branches works when the Court reads a remedial statute more narrowly than its authors intended.

Understanding the cycle matters for reading the 2008 act correctly. An override is not a repudiation of the judiciary’s role. It is the system functioning as designed: courts interpret ambiguous text, Congress clarifies the text when it disagrees with the interpretation, and the clarified text governs going forward. The 2008 act’s unusual explicitness, the naming of decisions in the findings, the point-by-point rejections in the purposes, reflects a Congress that had watched nine years of litigation and wanted no ambiguity about what came next. The clarity worked. The post-2009 case law on the definition is a fraction of what came before, not because the questions vanished but because the statute now answers them.

For readers who want to trace what changed in practice after the definition reopened, the effects are visible in charge processing, in the kinds of cases that reach trial, and in the advice employment counsel now gives. The through line is the one this article has followed from the start. Congress amended the statute for the express purpose of overturning Supreme Court decisions it said had read disability too narrowly. It named the holdings, rejected them in the purposes, and replaced them with four mechanical changes and a restructured third prong. Litigation moved from whether the plaintiff is disabled to whether the accommodation was reasonable. The duty never changed. The doorway did. That is the whole story, and the statute tells it on its face.

Why the override endures

Override statutes face a familiar risk: courts distinguish them, narrow them, and gradually restore the old regime under new reasoning. The 2008 act was drafted against that risk with unusual self-awareness. Its defenses are layered. The purposes name the rejected reasoning, so a court cannot claim ignorance of what Congress meant to displace. The Code provisions write the replacement rules with illustrative lists, so a court cannot plead ambiguity about what replaces the old holdings. The broad-construction mandate and the directive against extensive analysis set the interpretive attitude, so a court cannot revive the demanding standard as a background principle. And the conforming amendments to the Rehabilitation Act prevent the old definition from surviving in a parallel forum. Each layer answers a different method of judicial narrowing, and together they form a structure that has held.

The deeper reason for the act’s durability is political rather than textual. The negotiated compromise gave both sides a stake in the new regime. Disability advocates defend the four reversals because they restored the intended coverage. Employer organizations defend the boundaries, the eyeglasses exception, the transitory-and-minor rule, the regarded-as accommodation limit, because those provisions made the restoration administrable. A future Congress tempted to revisit the definition would face a coalition on both sides of the current text, each protecting its gains. Overrides last when they become settlements, and the 2008 act became one before it was even signed, in the months of bargaining that produced the lopsided votes. The statute endures not only because it is well drafted, though it is, but because the process that produced it left no major stakeholder wanting it reopened.

A map of the amended definition

The amended 42 U.S.C. 12102 can be read as a map, and the reader who has followed this guide now holds the legend. Section 12102(1) states the three prongs: actual disability, record of disability, and regarded as. Section 12102(2) defines major life activities through the two non-exhaustive lists, the outward activities and the major bodily functions. Section 12102(3) rebuilds the regarded-as prong around the actual-or-perceived impairment test, fenced by the transitory-and-minor exception. Section 12102(4) states the rules of construction: broad coverage, consistency with the findings and purposes, the episodic rule, and the mitigating-measures disregard with its eyeglasses exception. Section 12201(h) adds the regarded-as accommodation bar.

Each element of the map answers a holding. The mitigating-measures disregard answers Sutton, Murphy, and Albertson’s. The broad-construction command answers Toyota. The regarded-as revision answers Sutton’s third-prong reasoning and restores Arline. The bodily-functions list answers the vocabulary gap that had stranded plaintiffs with internal impairments. The episodic rule answers the temporal gap that had stranded plaintiffs with variable conditions. The override table near the beginning of this guide pairs each answer with its question; the map in this closing section shows how the answers fit together into a single coherent definition.

The statute that emerges from the map is the statute Congress said it had enacted in 1990: a broad definition opening onto unchanged duties, unchanged defenses, and unchanged remedies. The battleground moved, the weapons stayed the same, and the litigation took up the questions the statute had always meant it to decide.

Researching the statute

A researcher working with the 2008 enactment needs four documents within reach: the text of Public Law 110-325, the findings and purposes in section 2, the amended definition at 42 U.S.C. 12102, and the EEOC’s implementing regulations at 29 C.F.R. part 1630 with the Commission’s questions and answers on the final rule. The findings supply Congress’s account of the narrowing; the purposes supply the repudiation; the amended definition supplies the operative commands; the regulations supply the agency’s translation, including the “not meant to be a demanding standard” formulation that careful writers attribute to the EEOC rather than to the statute.

Two cautions should guide the research. First, always check the date of the underlying conduct before applying the definition: the January 1, 2009 boundary divides the old doctrine from the new, and opinions decided after the boundary may still apply the old definition to earlier facts. Second, keep the attribution of each proposition straight: the findings describe, the purposes reject, the subsections command, and the regulations gloss. Blurring those layers is the most common error in secondary writing about the statute, and the statute’s own structure is the cure.

The findings as a legislative record

Section 2(a)’s findings deserve a final look, because they do unusual work for findings. Most statutory findings are hortatory, broad statements of purpose that courts cite for background and little else. The 2008 findings are different. Paragraph (4) names Sutton, its companion cases, and Toyota, and states that their holdings narrowed the intended scope of protection. The naming converts the findings from background into evidence: a contemporaneous congressional statement, enacted as law, about what specific decisions did and why Congress disagreed. A court interpreting the amended definition cannot treat the override as an inference or a matter of legislative history buried in committee reports. It is in the statute, in the section the court must read before reaching the operative provisions.

The findings also serve a memorial function. They preserve, in the Code itself, the story of the nine-year narrowing: what the Court held, what Congress thought of the holdings, and what the statute was meant to do instead. Future readers who encounter the amended definition without knowing the trilogy will find the explanation in the findings that precede it. That self-contained quality is rare in legislation and deliberate here. Congress was writing for an audience that included future courts tempted to narrow the definition again, and it left those courts a record that makes the temptation harder to indulge. The findings say, in effect: we saw what happened last time, and we are telling you what we meant.

This is why the attribution discipline matters, the rule that the rejections live in the purposes rather than the findings. The findings describe the problem. The purposes state the solution. A reader who conflates the two will misattribute the operative language and weaken the citation. But a reader who understands the division of labor will see the section for what it is: a complete theory of the statute, problem, solution, and instruction, stated in the order a court needs it. Few sections of the United States Code are so candid about their own purpose. The candor is the point.

What makes the 2008 findings unusual?

Most statutory findings offer broad background, but section 2(a) names the specific Supreme Court decisions Congress believed narrowed the statute and states how they did it, converting the findings into an enacted record of disagreement that courts must confront when interpreting the amended definition.

Frequently Asked Questions

Q: What did the ADA Amendments Act change?

The act rewrote the definition of disability to reverse Supreme Court decisions that had narrowed coverage. It made four mechanical changes: mitigating measures other than ordinary eyeglasses are disregarded, “substantially limits” is construed broadly rather than as a demanding standard, episodic or in-remission impairments qualify if limiting when active, and major life activities include major bodily functions. It also restructured the regarded-as prong so plaintiffs need only show an actual or perceived impairment. The accommodation duty, qualification requirement, defenses, and remedies were left untouched.

Q: Which Supreme Court decisions did the ADA Amendments Act overturn?

Congress targeted four decisions. The 1999 trilogy, Sutton v. United Air Lines (527 U.S. 471), Murphy v. United Parcel Service (527 U.S. 516), and Albertson’s v. Kirkingburg (527 U.S. 555), all decided June 22, 1999, held that mitigating measures must be considered when assessing disability. Toyota Motor Manufacturing v. Williams (534 U.S. 184, 2002) construed “substantially limits” as a demanding standard. Section 2(a)(4) findings name these holdings as narrowing the statute’s intended scope, and section 2(b) purposes expressly reject them.

Q: Do mitigating measures count under the ADA Amendments Act?

No, with one exception. Under 42 U.S.C. 12102(4)(E), the disability determination is made without regard to the ameliorative effects of mitigating measures, including medication, prosthetics, hearing aids, assistive technology, and learned behavioral or adaptive neurological modifications. The sole exception is ordinary eyeglasses and contact lenses, whose corrective effects are still considered. This reverses Sutton, Murphy, and Albertson’s, which had required courts to account for mitigation, including the body’s own unconscious adaptations. The reversal is complete: courts may no longer treat effective treatment as evidence that no disability exists.

Q: What are major bodily functions in the ADA Amendments Act?

Under 42 U.S.C. 12102(2)(B), major life activities include the operation of major bodily functions. The statute gives a non-exhaustive list: immune system, normal cell growth, digestive, bowel, bladder, neurological, brain, respiratory, circulatory, endocrine, and reproductive functions. Because the list is introduced by “including,” courts may recognize additional functions. The provision lets conditions like cancer (normal cell growth), diabetes (endocrine function), and HIV (immune function) qualify directly without showing limits on traditional activities like walking or seeing.

Q: Does the regarded as prong still work under the ADA Amendments Act?

Yes, and it works better for plaintiffs than before. Under 42 U.S.C. 12102(3)(A), a plaintiff satisfies the prong by showing she faced a prohibited action because of an actual or perceived impairment, whether or not the impairment limits a major life activity. Two limits apply: impairments that are both transitory (six months or less) and minor are excepted under 42 U.S.C. 12102(3)(B), and regarded-as-only plaintiffs are not entitled to reasonable accommodation under 42 U.S.C. 12201(h).

Q: Did the ADA Amendments Act make it easier to prove disability?

Yes, deliberately. Congress directed that “substantially limits” be construed broadly in favor of coverage under 42 U.S.C. 12102(4)(A), and section 2(b)(5) states the definition should not demand extensive analysis. The EEOC’s regulations add that the standard is not meant to be a demanding one (29 C.F.R. 1630.2(j)(1)(i)). Combined with the mitigating-measures, episodic-condition, and bodily-function rules, these changes mean far fewer plaintiffs lose at the threshold, and litigation now centers on qualification and reasonable accommodation instead.

Q: When did the ADA Amendments Act take effect?

The act became effective January 1, 2009, under its own section 8. President George W. Bush signed it September 25, 2008, as Public Law 110-325. The EEOC takes the position that the amendments are not retroactive: discriminatory acts occurring before January 1, 2009 are judged under the pre-amendment definition shaped by Sutton and Toyota, even if the charge was filed later, while acts on or after that date are judged under the new text. The gap between the September signing and the January effective date gave employers and the agency time to prepare for the new framework.

Q: Was the ADA Amendments Act bipartisan?

Yes, emphatically. The Senate passed S. 3406 by unanimous consent on September 11, 2008, and the House passed the Senate bill by voice vote on September 17, 2008. The House had earlier passed its own H.R. 3195 by 402 to 17 on June 25, 2008. President George W. Bush, a Republican, signed it, eighteen years after his father, President George H. W. Bush, signed the original 1990 act. Business groups and disability advocates both supported the final negotiated text.

Q: What public law number carries the ADA Amendments Act?

The enactment is Public Law 110-325, recorded at 122 Stat. 3553. The public law number identifies the statute in the session laws of the 110th Congress: the 325th public law enacted during that Congress. Researchers citing the statute’s findings and purposes, including the rejection of the Sutton trilogy and Toyota, should cite the public law for the enacted text and 42 U.S.C. 12102 for the amended definition as codified. The Statutes at Large citation, 122 Stat. 3553, gives the page in the official compilation where the enrolled text appears. Together the two citations locate the statute in both the chronological and the codified collections of federal law.

Q: Which bill became the ADA Amendments Act?

The enacted bill was S. 3406, introduced in the Senate by Senator Tom Harkin. The Senate adopted it by unanimous consent on September 11, 2008, and the House passed the Senate bill by voice vote on September 17, 2008. The House had earlier passed its own bill, H.R. 3195, by 402 to 17 on June 25, 2008, but the House set that bill aside and took up the Senate version, so the June roll call is not the final House action. President Bush signed S. 3406 on September 25, 2008, and it became Public Law 110-325. Legislative-history researchers should work from the Senate bill’s reports and floor statements, not the House bill’s, when tracing the enacted text.

Q: Why is the 2008 signing described as a bipartisan bookend?

Because the two signatures eighteen years apart came from father and son of different parties’ eras uniting on the same cause. President George H. W. Bush signed the original Americans with Disabilities Act on July 26, 1990, and his son, President George W. Bush, signed the 2008 amendments restoring its definition on September 25, 2008. Combined with the unanimous-consent Senate vote and the voice-vote House passage, the family symmetry made the amendments a symbol of disability policy as shared national commitment rather than partisan project.

Q: Where did the phrase not meant to be a demanding standard come from?

The phrase comes from the EEOC’s implementing regulations, not from the statute. It appears at 29 C.F.R. 1630.2(j)(1)(i), in the Commission’s rules construing the phrase “substantially limits.” The statute itself, at 42 U.S.C. 12102(4)(A) through (B), commands broad construction in favor of expansive coverage and interpretation consistent with the findings and purposes, and the purposes in section 2(b)(4) expressly reject the demanding standard Toyota had announced. The regulatory phrase faithfully renders that statutory command, but careful writers attribute it to the EEOC. The distinction matters because it keeps straight which branch said what: Congress rejected the standard, and the agency supplied the memorable formulation.

Q: Why are ordinary eyeglasses treated differently under the ADA Amendments Act?

The eyeglasses exception was the product of legislative bargaining. Section 12102(4)(E) disregards the ameliorative effects of every listed mitigating measure except ordinary eyeglasses and contact lenses, whose effects are still considered. The exception preserves the result of Sutton v. United Air Lines on its own facts, the rejected pilot applicants whose vision was fully corrected by lenses, while discarding the rule Sutton announced for everything else. It gave the business community one stable, familiar rule amid a statute that changed nearly everything else about coverage, and it drew a clean line between routine vision correction and every other form of medical management, including low-vision devices, which fall on the disregarded side.

Q: How does the ADA Amendments Act treat impairments that are episodic or in remission?

Section 12102(4)(D) provides that an impairment that is episodic or in remission is a disability if it would substantially limit a major life activity when active. The court fixes the temporal frame on the active phase rather than on the quiescent moment of measurement. Epilepsy between seizures, asthma between attacks, depression between episodes, cancer in remission, and multiple sclerosis in its relapsing course are the familiar illustrations, though the statute names no conditions. The provision works together with the mitigating-measures rule: a medicated episodic condition presents both the ameliorative effect of the drug and the quiescence of the condition, and the statute disregards both in the coverage analysis.

Q: What is the transitory and minor exception in the regarded as prong?

Section 12102(3)(B) excepts from regarded-as coverage any impairment that is transitory and minor. Transitory means an actual or expected duration of six months or less, and the exception is conjunctive: the impairment must be both transitory and minor to fall outside the prong. A transitory impairment that is not minor, such as a serious injury expected to heal within five months, remains covered, as does a minor impairment expected to last longer than six months. The exception is the remaining boundary of the rebuilt prong, replacing the demanding showing Sutton had required. Plaintiffs invoking the prong should be prepared to address both halves of the conjunction, duration and severity.

Q: Can an employer deny a reasonable accommodation to someone covered only as regarded as disabled?

Yes, the statute permits it. Under 42 U.S.C. 12201(h), covered entities need not provide reasonable accommodations or modifications to individuals whose only basis for coverage is the regarded-as prong. The logic is that the prong targets stigma and mistaken perception rather than functional limitation, so the remedy is to stop the discriminatory act, not to restructure the job. A regarded-as plaintiff who also satisfies the actual-disability prong retains full accommodation rights. The distinction keeps the prong a remedy for stereotyping rather than a source of workplace restructuring claims.

Q: What earlier case did Congress restore for the regarded as prong?

Congress restored the reasoning of School Board of Nassau County v. Arline, 480 U.S. 273 (1987), a pre-Sutton decision that took a broad view of regarded-as coverage. Section 2(b)(3)’s purposes expressly reject Sutton’s narrowing of the third prong and reinstate Arline’s approach, under which an individual subjected to a prohibited action because of others’ reactions to an impairment is covered. The restored rule is codified at 42 U.S.C. 12102(3)(A), requiring only an actual or perceived impairment behind the employer’s action.

Q: Did the ADA Amendments Act change the Rehabilitation Act?

Yes, by conforming amendment. Section 7 of the 2008 enactment amends the definition section of the Rehabilitation Act of 1973 so that the two statutes share a single disability definition. The amendment operates at two points in 29 U.S.C. 705, paragraph (9)(B) and paragraph (20)(B), conforming both to the definition in section 3 of the Americans with Disabilities Act as amended. The result is that federal-sector employment, federally funded programs, and the private-sector titles of the ADA all apply the same restored definition, with the same disregarded mitigating measures, the same broad construction, the same episodic rule, and the same bodily-functions list. The two statutes had long been read in tandem, and Congress wrote the tandem reading into the text.

Q: Does the ADA Amendments Act apply to conduct before 2009?

No. The statute is not retroactive. Its section 8 sets the effective date at January 1, 2009, and the EEOC’s questions and answers on the final implementing rule confirm that the revised definition governs only alleged discriminatory acts occurring on or after that date. A termination, a failure to hire, or a denied accommodation that occurred in December 2008 is assessed under the original definition, with the Sutton mitigating-measures rule and the Toyota demanding standard intact, even if the worker did not file a charge until after the new year. Opinions handed down in 2009 or later may therefore still apply the old definition when the underlying facts predate the boundary.

Q: Where did disability litigation move after the definition reopened?

It moved from the threshold to the merits. Before 2009, employers most often won by arguing the plaintiff was not disabled, and courts dismissed cases before reaching the employer’s conduct. After the amendments, that defense largely stopped working, so disputes now center on whether the person is qualified to perform the job’s essential functions and whether the requested accommodation was reasonable. Plaintiffs’ lawyers spend less on threshold medical experts; defense counsel litigate the interactive process, undue hardship, and essential functions instead.