Every medicine sold in India, from a strip of paracetamol at a village chemist to a patented oncology injection in a metropolitan hospital, passes through a single statutory doorway: the Drugs and Cosmetics Act 1940. The Act decides what may be imported, what may be manufactured, what may be sold, by whom, under what conditions, and with what consequences when the product on the shelf is not what its label claims. Yet the statute most readers imagine, a single regulator guarding a single gate, does not exist. In its place stands a divided machine. A new medicine is approved once, at the Centre, by the Central Drugs Standard Control Organisation headed by the Drugs Controller General of India. The licence to manufacture that same medicine is then granted state by state, and the licence to sell it is granted state by state again. Approval is national; permission to make and permission to sell are local. Almost everything that is distinctive about Indian drug regulation, its reach, its unevenness, and its recurring quality scandals, follows from that one structural fact.

The claim this article defends, and returns to at each stage, can be stated in one line: a medicine in India is approved once and licensed thirty-six times, so the quality of what reaches a patient depends less on the approval the product received than on the capacity of the licensing authority in the state where it was made. Twenty-eight states and eight union territories each run a licensing jurisdiction under the Act. The Centre writes the standards, approves the new molecule, and controls the border. The states license the factory and the shop. When a batch fails, the failure is almost always located in the second half of that sentence. This article is the statute pillar for the drug-regulation cluster within the wider account of health legislation in India, and it takes the Act apart in the order a reader needs it: what came before, how the text is organised, the definitions that decide prosecutions, the institutions, the approval machine, the licensing machine, the schedules where practice actually lives, the penalty ladder, the courts, the two great official indictments of the system, and the replacement statute that has been drafted and not enacted.

Before the Act: an unregulated market

The statute was born of a market failure documented a decade before independence. In the first decades of the twentieth century India had no general law controlling the quality of medicines manufactured in the country or imported into it. The Indian Penal Code punished adulteration of drugs as a public-health offence, and the Opium Act and the Dangerous Drugs Act controlled narcotics, but no licensing system stood between a manufacturer and the market. Imported medicines arrived from Britain, Germany and elsewhere with no Indian standard they had to meet, and domestic production ranged from the laboratories of established firms to back-room operations whose products bore impressive labels and uncertain contents.

The pressure for legislation built through the 1920s. A resolution in the Council of States in 1927 pressed the Government of India to control the preparation and sale of medicinal drugs, and in 1930 the government appointed the Drugs Enquiry Committee, chaired by Lieutenant Colonel R. N. Chopra, to examine the extent of adulteration and spurious manufacture and to recommend a system of control. The Chopra Committee reported in 1931. It found a market flooded with adulterated and spurious products, identified the absence of any central drug control organisation and of adequate testing laboratories as the root defects, and recommended comprehensive central legislation on the lines of the British and American statutes: control of imports, licensing of manufacture and sale, government analysts and testing laboratories, and an expert advisory board to keep the law abreast of pharmaceutical science.

The demand had a documentary trail before the committee sat. The 1927 resolution had been moved in the Council of States by Sir Haroon Jaffer, and the Government of India’s response, a letter of April 1927 to the local governments, asked each province to examine the formation of machinery to control the manufacture and sale of drugs and the standards such machinery should enforce. The replies exposed the patchwork in official prose: provinces with a poisons law and little else, ports through which imported medicines passed untested, and no laboratory anywhere in the system charged with saying what a medicine actually contained. Contemporary investigation filled in the human detail; studies published in the Indian Medical Gazette in the late 1920s on stock solutions of quinine, the empire’s staple antimalarial, found preparations in the market whose strength bore little relation to their labels. The committee’s achievement was to convert that scattered record into a single institutional design, and the statute of 1940 is that design enacted almost whole.

Legislation took most of a decade to arrive, and its route explains its federal shape. Under the Government of India Act 1935, drugs fell within the provincial legislative field, so a central statute required the provinces to pass resolutions under Section 103 of that Act authorising Parliament at the Centre to legislate on their behalf. The preamble to the 1940 Act records exactly that: the legislatures of all the provinces passed the necessary resolutions. The Drugs Act, as it was first called, received the assent of the Governor General on 10 April 1940 as Act 23 of 1940. Cosmetics were brought within its reach by the amendment of 1962, which inserted “and Cosmetics” into the title with effect from 27 July 1964 and supplied the definition of a cosmetic and the chapter machinery to govern it.

Commencement was staged, and the staging matters to any lawyer reading an old case. Section 1(3) brought the Act into force at once in formal terms, but Chapter III, which governs imports, took effect only from a date the Central Government appointed, which was 1 April 1947, and Chapter IV, which governs manufacture, sale and distribution, took effect in each state only from a date that state government appointed. Delhi, Ajmer and Coorg applied Chapter IV from 1 April 1947; other territories followed in later years, some as late as 1953 and beyond. The Drugs and Cosmetics Rules, framed in 1945, supplied the operative detail in advance of the chapters they served. A statute whose operative parts switched on province by province across six years was, from birth, a national text administered at provincial speed. That character never left it.

What grew on this grid after independence is the fact every later controversy presupposes. The licensed factory system, the pharmacopoeial standard and the state inspectorates of the 1940 settlement became the platform on which India built, over the following half century, the largest generic medicines industry in the world by volume, the industry the government routinely describes as the pharmacy of the world for the share of global generic and vaccine supply it carries. The growth was not the Act’s doing alone; it belongs equally to the patent settlement of 1970 and to the price orders traced later in this article, and the account of the health laws that frame this field, the companion guide in this series, places the three in their order. But the platform mattered: an exporter selling into the regulated markets of the West was, at home, a licensee of a state authority under a central code, and every weakness of that home regulation, every unfilled analyst post and uninspected factory, is therefore also a fact about the credibility of a global supply chain. The Act of 1940 was written for a colony importing finished medicines. It underwrites in this century an export industry its drafters could not have pictured, which is the deepest reason its replacement is debated as industrial policy and not merely as health law.

The shape of the Drugs and Cosmetics Act 1940 and the Rules beneath it

The parent Act is short, and its brevity is the first thing a practitioner learns about it. Its chapters divide the field by point of control. Chapter I contains the short title, extent and the definitions in Section 3. Chapter II creates the advisory and laboratory machinery: the Drugs Technical Advisory Board under Section 5, the Central Drugs Laboratory under Section 6 and the Drugs Consultative Committee under Section 7. Chapter III governs import, with its own set of offence definitions and its own licensing of what crosses the border. Chapter IV governs manufacture, sale and distribution within India, and carries the definitions, prohibitions and penalties around which this article is built. Chapter IVA, inserted in 1964, is a self-contained code for Ayurvedic, Siddha and Unani medicines. Chapter V is miscellaneous: offences by companies, confiscation, jurisdiction, the rule-making powers in Sections 12 and 33, and the protection for action taken in good faith.

Two features of this architecture deserve emphasis. The first is what the Act does not contain. It prescribes almost no standards itself. The standard a medicine must meet is located by Section 16 in the Second Schedule to the Act, which points outward to the Indian Pharmacopoeia and, for products the Pharmacopoeia does not cover, to other official pharmacopoeias and formularies named there. The good manufacturing practice a factory must observe is not in the Act at all; it is Schedule M to the Rules. The categories of medicine a chemist may sell only against a prescription are not in the Act; they are Schedules H, H1 and X to the Rules. The conditions of every licence, the forms of every application, the fees, the qualifications of the competent technical staff who must supervise manufacture, the shelf life permitted to each formulation: all of it sits in the Drugs and Cosmetics Rules 1945, a subordinate instrument many times the length of its parent. Whoever asks what Indian drug law requires, on almost any concrete question, is asking a question about the Rules and their schedules, and only through them about the Act.

The second feature is the constraint on that rule-making power, which is real and frequently forgotten. Sections 12 and 33 oblige the Central Government, before making rules under the import and manufacture chapters respectively, to consult the Drugs Technical Advisory Board, and to publish draft rules in advance for objections and suggestions. The consultation requirement is a condition of validity, not a courtesy: rules made without the Board’s consultation are vulnerable in court. The statute thereby hard-wires expert consultation into its own expansion, which is one reason the law in this field develops by notified amendment and committee recommendation rather than by fresh legislation. The series thesis for this cluster takes its sharpest form here. A statute whose substance lives in schedules and rules, whose parent text has been overtaken by a drafted replacement that Parliament has not taken up, is governed in practice by notifications. The sections that follow test that proposition against each part of the machine.

Three exhibits state the proposition before the test begins. In February 2020 a single notification under Section 3(b)(iv) multiplied the regulated universe severalfold by declaring every medical device in the country a drug, a change in the industrial landscape larger than most Acts of Parliament achieve. In March 2019 a single set of rules replaced the clinical trial code wholesale, timelines, compensation formula and all, without a clause of the parent Act moving. In December 2023 a single gazette notification rewrote the good manufacturing practice schedule against which every factory in India is judged. In each case the operative law of Indian drug regulation changed completely while the statute book, consulted the way a citizen consults it, stood still, and in each case the change was lawful, because the Act had from the beginning delegated its substance to the instruments beneath it. The reader who wants the law of Indian medicines on any given date must therefore read the gazette alongside the statute, or accept that the statute alone will deceive by omission. That is not a complaint about delegation, which a technical field cannot do without; it is a statement about where the constitution of this field actually sits, in the consultation duty owed to the Board, the publication of draft rules, and the laying of rules before Parliament, the procedural safeguards that substitute for the parliamentary debate the parent Act will never again receive on its substance. A pillar article on this statute is therefore necessarily an article about two texts: the short, stable, much-amended Act, and the long, restless, notification-fed code beneath it that practitioners actually obey.

The definitions in Section 3 set the perimeter. A “drug” under Section 3(b) includes all medicines for internal or external use of human beings or animals and all substances intended for use in the diagnosis, treatment, mitigation or prevention of any disease or disorder; substances intended to affect the structure or any function of the human body, or to be used for the destruction of vermin or insects which cause disease; substances intended for use as components of a drug, including empty gelatin capsules; and such devices intended for internal or external use in diagnosis, treatment, mitigation or prevention of disease as the Central Government specifies by notification after consulting the Board. The breadth is deliberate. Because components and capsules are inside the definition, the supplier of an ingredient or a shell is inside the licensing net. Because devices enter only by notification, the device sector’s position under the statute has always depended on the notification history traced later in this article. A “cosmetic” under Section 3(aaa) is any article intended to be rubbed, poured, sprinkled or sprayed on, or introduced into, or otherwise applied to the human body for cleansing, beautifying, promoting attractiveness or altering appearance, and includes any article intended for use as a component of a cosmetic. Two working consequences follow. A product marketed with a therapeutic claim is a drug however cosmetic its presentation, because the definition follows intended use; and a product making no therapeutic claim may escape the drug definition entirely and answer only to the cosmetics provisions, which is why the borderline between the two categories is litigated with such energy by manufacturers of everything from skin creams to herbal preparations.

One caution belongs in every account of this statute, and it is given once here. What follows is a general account of the law for readers who need to understand the system; it is not advice on any particular licence, product or prosecution, and anyone facing a notice, a prosecution or a licensing decision under the Act needs advice on the facts of that case, because limitation periods, sampling timelines and forum rules in this field are unforgiving.

How the Act grew: the amendments that made the modern statute

The text of 1940 would be barely recognisable to its drafters in five respects, and the amendment history is the shortest route to understanding which parts of the machine are original and which are accretions. The amendment of 1962 brought cosmetics into the statute and its title, converting a drugs Act into the joint code it has remained. The amendment of 1964 inserted Chapter IVA and thereby acknowledged that Ayurvedic, Siddha and Unani medicines could not be governed sensibly by provisions written for synthetic pharmaceuticals, an early admission that one statute would have to house several regulatory logics. The amendment of 1972 completed the statute’s territorial extent by removing the exclusion of Jammu and Kashmir from its reach.

The amendment of 1982, Act 68 of 1982, in force from 1 February 1983, was the first great strengthening. It substituted Sections 27 and 27A with a graduated penalty ladder in which a spurious or adulterated drug likely to cause death or grievous hurt drew imprisonment from five years to life, and a spurious drug drew three to five years, figures that look modest beside their successors but transformed the offence from a fining matter into a serious crime. It inserted Section 26A, the Central Government’s power to prohibit a drug in the public interest, which has since become the most politically visible provision in the Act, and Section 28A, penalising the failure to keep documents and furnish information to inspectors. It also tightened the definitional scheme around the adulterated and spurious categories. The direction of travel was set in 1982 and has never reversed: each succeeding amendment has raised the penal temperature while leaving the licensing geography untouched.

The amendment of 2008, Act 26 of 2008, in force from 10 August 2009, was the Mashelkar amendment, and its changes have been threaded through this article: the Section 27 ladder raised to ten years to life for the harm-causing offences with fines measured in lakhs and directed to the victim as compensation, Section 17E added to complete the cosmetics definitions, offences under Chapter IV made triable only by a Court of Session, Special Courts authorised under Section 36AB, the listed adulterated and spurious offences made cognizable with restricted bail under Section 36AC, and compounding of minor offences introduced by Section 32B. Read against the amendment of 1982, the 2008 Act is the same statute thinking harder about the same problem: the definitions and the licensing split stand as they stood, and what changes is the severity, speed and specialisation of the process brought against the counterfeiter. The Jan Vishwas amendments of 2023, which trimmed the periphery of that severity for minor regulatory offences, are the first amendment in the sequence to move in the opposite direction, and their narrowness, three sections touched, is the measure of how settled the penal core has become.

The definitions that decide prosecutions

Chapter IV does not punish bad medicines in the abstract. It punishes them through four constructed categories, each defined by a separate section, each carrying a different consequence, and the whole of drug prosecution practice consists of assigning a failed product to the right category. A prosecutor who charges the wrong one loses. A report that uses the four words interchangeably, as press coverage habitually does, tells its reader nothing about the case it describes.

Standard quality is the baseline, defined by Section 16. For the purposes of Chapter IV, a drug is of standard quality if it complies with the standard set out in the Second Schedule, which in practice means the standard of the Indian Pharmacopoeia for products the Pharmacopoeia covers. A sample that fails the pharmacopoeial test, because its active ingredient assays below the permitted limit, or it fails a dissolution or uniformity requirement, or its shelf life has expired, is described throughout the system as “not of standard quality”, universally shortened to NSQ. NSQ is a conclusion about quality, not about identity or intent. It covers the honestly made tablet that degraded in a hot godown and the carelessly made tablet that left the factory weak. The category is the workhorse of enforcement: the great majority of failed samples in every official survey are NSQ, and the legal system’s treatment of an NSQ finding, administrative action for minor deviations, prosecution for gross ones, reflects that spread.

A drug is misbranded under Section 17 if it is so coloured, coated, powdered or polished that damage is concealed, or it is made to appear of better or greater therapeutic value than it really is; if it is not labelled in the prescribed manner; or if its label, container or anything accompanying it bears a statement, design or device making a false claim for the drug or which is false or misleading in any particular. Misbranding is a sin of presentation. The contents may be perfect; the offence is in what the packet says or hides. It is the least grave of the three named categories, and the one most often engaged by labelling prosecutions.

A drug is adulterated under Section 17A if it consists in whole or in part of any filthy, putrid or decomposed substance; if it has been prepared, packed or stored under insanitary conditions whereby it may have been contaminated with filth or rendered injurious to health; if its container is composed of a poisonous or deleterious substance which may render the contents injurious; if it bears a colour other than a prescribed colour; if it contains any harmful or toxic substance which may render it injurious to health; or if any substance has been mixed with it so as to reduce its quality or strength. Adulteration is a sin of contamination or debasement: something is in the product that should not be there, or the conditions of its making have poisoned it, or a cheaper substance has been folded in to stretch it. The emphasis falls on danger to the body rather than deception of the eye.

A drug is spurious under Section 17B if it falls within any of five limbs, every one of which is built on false identity. It is spurious if it is manufactured under a name which belongs to another drug; if it is an imitation of, or a substitute for, another drug, or resembles another drug in a manner likely to deceive, unless it is plainly and conspicuously marked to reveal its true character; if the label or container bears the name of an individual or company purporting to be the manufacturer which is fictitious or does not exist; if it has been substituted wholly or in part by another drug or substance; or if it purports to be the product of a manufacturer who did not in fact make it. The spurious drug is, in substance, the counterfeit: a product lying about what it is or who made it. Nothing in Section 17B requires the contents to be weak or dangerous. A spurious tablet of full pharmacopoeial strength is still spurious, because the offence is the deception. Conversely, a genuine manufacturer’s tablet that assays at half strength is not spurious at all; it is NSQ, and may be adulterated if the shortfall came from dilution, but the spurious charge fails for want of any false identity.

That distinction, elementary as it is, decides real cases, and it is worth stating the analytical rule plainly because enforcement reporting so often blurs it. A failed assay, without more, establishes at most a drug not of standard quality. A spurious prosecution must prove one of the Section 17B limbs: imitation, substitution, a fictitious maker, or a false claim of manufacture. Where the two overlap, as they do when a counterfeiter fills genuine-looking strips with chalk, both charges lie, and the penalty ladder examined later reserves its highest rungs for the combination. The categories for cosmetics run parallel but not identical: Section 17C defines a misbranded cosmetic, Section 17D a spurious cosmetic, and Section 17E, inserted by the amendment of 2008, an adulterated cosmetic, and Chapter III carries mirror definitions for imports in Sections 9 to 9D, so that the same consignment can be characterised at the border and in the market by the same conceptual scheme.

The mirror, however, reflects the definitions and not the consequences, and the asymmetry is one of the strangest survivals in the Act. Section 13 punishes the import of an adulterated or spurious drug with imprisonment extending to three years and a fine extending to five thousand rupees, and a second conviction with imprisonment extending to five years or a fine extending to ten thousand rupees. The amendment of 2008, which rebuilt the domestic penalty ladder almost from the ground, left Chapter III essentially untouched, so the importer of a container of counterfeit medicine faces, on the face of the statute, fines fixed at the price level of the 1940s, while the domestic seller of a single strip of the same product faces the modern ladder. In practice the gap is papered over: import consignments are stopped administratively at the port, licences and registration certificates are suspended or cancelled, and the goods are confiscated under Section 14 or re-exported or destroyed under the customs machinery. But the textual survival matters to the series thesis. A statute amended episodically, chapter by chapter and crisis by crisis, accumulates strata from different decades that no single drafting intelligence has reconciled, and Chapter III is the stratum from 1940 still showing through.

The three domestic labels are easiest to keep straight on a single shelf, and a worked tour of one chemist’s stock shows how the scheme classifies in practice. A strip of a well-known antacid, made by its genuine manufacturer, in date, correctly labelled, but assaying at eighty per cent of the declared strength because a batch was badly blended, is a drug not of standard quality and nothing worse: the fraud element is absent, the failure is one of manufacture, and the prosecution, if any, runs on the analyst’s report alone. The same manufacturer’s genuine strip, past its expiry date and relabelled by a distributor with a fresh date, is misbranded, because the relabelled date deceives, and it may also be not of standard quality, since an expired drug that has degraded fails the pharmacopoeial test too; the two labels accumulate on one strip of tablets. A strip printed to imitate the genuine product, containing chalk and a trace of the named drug, made by an unlicensed unit and passed into the wholesale chain with forged invoices, is spurious, and if its contents are likely to cause death or grievous hurt when taken as directed, it stands on the top rung of the ladder. A bottle of imported cough mixture whose colour has been loaded with a prohibited coal-tar dye is adulterated under the colour limb even if every other ingredient is genuine. A fairness cream bearing another company’s brand and get-up is a spurious cosmetic; the same cream with a misleading label claim is misbranded. The inspector’s craft is to state, for each seized article, which definition it answers and why, because the charge, the forum’s expectations and the sentence all follow the label, and a prosecution that pleads the wrong label for the right seizure invites the exact paperwork defeats the case law records.

The institutions the Act created, and the one it did not

Chapter II builds three bodies, and Indian drug administration is unintelligible until their distinct characters are fixed. The Drugs Technical Advisory Board, constituted under Section 5, is the statute’s expert brain. It consists of ex officio members drawn from the health administration, including the Director General of Health Services, who chairs it, and the Drugs Controller, together with nominated members representing pharmacology, medicine, pharmaceutical chemistry, the pharmaceutical industry and the pharmacy profession, and members elected by bodies such as the Medical Council and the Pharmacy Council. Its function is to advise the Central and state governments on technical matters arising out of the administration of the Act, and, as the rule-making sections show, its consultation is a statutory precondition to the making of rules. The Board recommends; it does not license, prosecute or decide. Its power is the power of the consulted expert whose concurrence the government must obtain before the law moves.

The Drugs Consultative Committee, constituted under Section 7, is the federal hinge. It consists of the Drugs Controller General as its chair and representatives of the licensing authorities of the states. Its statutory function is to advise the Central Government, the state governments and the Board on any matter tending to secure uniformity throughout India in the administration of the Act. Uniformity is the operative word. Because the Act is administered by more than three dozen licensing jurisdictions, the Committee exists precisely to hold the line against divergence: when a question arises whether a category of product should be treated one way in Gujarat and another in Odisha, the Committee is the forum in which a common answer is supposed to be forged. The recurring complaint of the official reviews, examined later in this article, is that it has forged common answers too rarely and enforced them hardly at all.

The Central Drugs Laboratory at Kolkata, established under Section 6 under a Director appointed by the Central Government, is the system’s court of final scientific appeal. State and central laboratories across the country test the samples that inspectors draw, and the Government Analysts appointed under Section 20 sign the reports on which prosecutions rest. But Section 25(4), examined with the enforcement machinery below, gives the report of the Director of the Central Drugs Laboratory a conclusive status no other laboratory report carries. Other institutes have been notified to discharge the Laboratory’s functions for particular classes of product, and a network of Central Drugs Testing Laboratories supports it, yet the statutory apex remains the Kolkata institution and its Director’s signature.

The laboratory system has two tiers, and the distance between them is where prosecutions are won and lost. At the base, each state maintains government analysts and a state drug testing laboratory, of widely varying capacity; some states historically possessed no full laboratory of their own and sent samples to laboratories in other states or to the central network. Above them stands the central tier: the Central Drugs Laboratory at Kolkata and the Central Drugs Testing Laboratories and regional laboratories maintained by the CDSCO in centres including Mumbai, Chennai, Hyderabad, Guwahati and Chandigarh, which test import samples, survey samples and the reserved portions sent under Section 25(4), and which conduct the batch testing of specified biological products, vaccines and sera among them, before release. The Government Analyst who signs a report is not a technician in the wings of the case; the report is evidence by statute, the analyst may be called and cross-examined, and the conclusive character the Act gives the report when uncontroverted makes the analyst’s bench the effective court of first instance for the question whether a medicine met its standard. The chronic weakness of the base tier, backlog, vacancies and equipment gaps in the state laboratories, propagates upward through everything the enforcement sections describe: a sample that waits a year for analysis is a prosecution launched against an expiring product, with the accused’s retest right decaying in the interval, which is precisely the fact pattern the Supreme Court’s sampling-delay jurisprudence addresses.

The body that dominates public discussion of drug regulation, the Central Drugs Standard Control Organisation, is conspicuously absent from this list, because the Act did not create it. The CDSCO is an administrative organisation of the Ministry of Health and Family Welfare, headed by the Drugs Controller General of India. Its legal authority arrives not from a constituting section but from the Rules, which designate the Drugs Controller General as the licensing authority for imports and for the approval of new drugs, and as the Central Licence Approving Authority whose concurrence is needed before a state may license certain sensitive categories of manufacture. The CDSCO works through zonal, sub-zonal and port offices, through which its officers clear imported consignments and conduct the central half of the joint inspections described later. Its position is therefore paradoxical and important: the most powerful regulator in the field is a creature of the executive, exercising statutory functions conferred by subordinate legislation, within an Act that gave its own three creations, a board, a committee and a laboratory, comparatively narrow and precise roles. Proposals to cure that anomaly by statute run through every reform document in this field, and the still-unenacted replacement bill would have given the central regulator a statutory foundation for the first time.

The constitutional footing of the whole arrangement explains why it takes this shape. Drugs and poisons stand in Entry 19 of the Concurrent List of the Seventh Schedule, so both Parliament and the state legislatures may legislate on the field, and the Constitution’s repugnancy rule in Article 254 gives the parliamentary law primacy where the two conflict. Parliament used that primacy to enact one code and one rulebook: the Act and the Central Rules bind every state, while the administration of the code, the licensing authorities, the inspectors, the analysts, is staffed and funded by the states under powers the Act itself distributes. A state may legislate on drugs for its own purposes, and several maintain cognate laws on cognate subjects, but a state law that contradicted the central code on the code’s own ground would yield to it. The settlement is therefore neither a Union subject administered by agents nor a state subject with central advice; it is a concurrent field in which the Centre wrote the text, the states work the text, and the friction between those two facts is not a defect of drafting but the field’s constitutional condition. Every reform proposal in this article, from Mashelkar’s central administration to the double key for manufacture, is an attempt to renegotiate that condition without amending the Constitution, which is why all of them proceed by rearranging functions under the existing Act rather than by redrawing the legislative map.

One office completes the institutional cast, and it is the one the citizen actually meets. The Act nowhere names the officer who signs a manufacturing or sale licence. The Rules supply the office: the licensing authority is the authority appointed by the State Government for the purpose, and in practice the function is exercised by the state’s Drugs Controller or Food and Drug Administration Commissioner through district and zonal licensing officers. The appointment matters because the licensing authority is a legal person in the system’s eyes: its orders are appealable, its show-cause notices begin the administrative track, its suspension of a licence stops a factory, and its name, not the Inspector’s and not the Drugs Controller General’s, is the name on the document without which no medicine may lawfully be made or sold in the territory. Students of the Act routinely misplace this office at the Centre because the Centre writes the rules the office applies; the signature, and with it the responsibility the federal debate argues about, is the state’s.

Below the Centre sits the state tier, which is where the Act touches the ground. Each state and union territory maintains a drug control organisation, headed typically by a Drugs Controller or a Commissioner of Food and Drug Administration, and licensing authorities appointed under the Rules who grant, renew, suspend and cancel the licences for manufacture and sale within the territory. The organisations are not uniform even in form: some states fold drug control into a combined food and drug administration, others run a separate drugs directorate, and the cadre strength, laboratory capacity and inspection tempo of the thirty-six jurisdictions diverge so widely that the state in which a factory happens to sit is, by itself, a material fact about the scrutiny that factory will receive. The Act prescribes one law for all of them and equips none of them; their budgets, posts and laboratories are creatures of state finance, and the distance between the best-equipped and the most skeletal state administrations is the single largest variable in the whole system. Sections 20 and 21 complete the cast: Government Analysts, appointed by the Centre and by the states, who test samples and report; and Inspectors, appointed by either government under Section 21, whose powers under Section 22 include inspecting any premises where a drug or cosmetic is manufactured or sold, taking samples, searching premises and seizing stock where the Act appears to have been contravened, and requiring records to be produced. The licensing pattern has siblings across the health cluster: the transplantation statute’s machinery of registered hospitals and authorisation committees rests on the same statutory instinct, permission first, supervision after, that governs a drug factory under this Act.

Approval at the Centre: new drugs and clinical trials

A medicine that has never before been marketed in India must clear a central gate before any state licence becomes relevant, and the law of that gate has been rewritten once, decisively, within the lifetime of most practising regulatory lawyers. Under the Drugs and Cosmetics Rules as they stood for decades, a “new drug” was defined by Rule 122E: broadly, a drug not previously used in the country to any significant extent, or a known drug proposed for a new indication, dosage form or route of administration, and such a drug remained a new drug for four years from its approval. Permission to import or manufacture it was governed by Rules 122A and 122B, and trials of it in India by Rule 122DA and Schedule Y, the long schedule of clinical trial requirements appended to the Rules in 1988. The structure worked, but it buried the system’s most consequential decisions in schedule text that had to be amended by notification for every refinement, and it left the ethics of trials to be assembled from Schedule Y, the Indian Council of Medical Research’s ethical guidelines and the licensing conditions.

The New Drugs and Clinical Trials Rules 2019, notified on 19 March 2019 and brought into force the same day for most of their chapters, lifted the whole subject out of Schedule Y and the old Part XA of the Rules into a free-standing code under the Act. The 2019 Rules define a new drug functionally, by reference to whether the drug has been used in the country to any significant extent and whether it is already approved and marketed for the claims proposed, rather than by the old four-year clock; the fixed period after which an approved product quietly ceased to be “new” disappeared from the definition, and with it a source of recurring argument about when central control of a product relaxed into ordinary state licensing. The Rules govern clinical trials, bioavailability and bioequivalence studies, the registration and conduct of ethics committees, the approval of new drugs for import and manufacture, and compensation for trial participants, and their animating trade-off, speed against protection, is argued in the compensation discussion that follows.

The approval decision is not made by the Drugs Controller General unaided, and the advisory layer deserves notice because the standing committee scandal of 2012, examined later, was at bottom a scandal about how expert advice was obtained. Applications for new drugs and trials are examined with the help of Subject Expert Committees, panels of clinicians and pharmacologists drawn from outside the regulator, whose recommendations the Authority ordinarily follows. The committees are administrative creations, not statutory bodies: the Act knows the Drugs Technical Advisory Board, and the Rules know the licensing authority, but the committee that in practice decides whether a molecule may be tested on Indian patients exists by office practice. That informality is defensible as flexibility and indictable as opacity, and the reforms after 2012, published committee compositions, declared conflicts of interest, recorded recommendations, were directed at exactly the gap between the committee’s real power and its paper status.

The trial permission machinery runs on stated timelines, a genuine novelty in Indian drug regulation. An application to conduct a clinical trial is made to the Central Licensing Authority, the Drugs Controller General, and the Rules commit the Authority to dispose of it within ninety working days. Where the drug is discovered in India, or is proposed to be researched and manufactured in India, the period is thirty working days, and if the Authority does not communicate within that period, permission is deemed to have been granted. Deemed approval on official silence is a strong medicine, and it was a deliberate choice: the sponsor community had argued for years that indeterminate waiting, rather than any stated requirement, was the true cost of doing research in India. The Authority may waive a local trial for a drug already approved in specified jurisdictions where the conditions the Rules prescribe are met, including the absence of any reason to expect a different response in the Indian population and a commitment to post-marketing surveillance in India; the waiver power is the point at which the question whether Indian patients are adequately represented in global data is decided, case by case, and it deserves the scrutiny it receives.

Ethics committees, the bodies at each trial site charged with protecting the participants rather than the science, are brought inside the statutory perimeter by the 2019 Rules. A committee that oversees a clinical trial must be registered with the Central Licensing Authority, its registration holds for five years, and the Rules prescribe its composition so that it cannot be a committee of the sponsor’s own people: it must include medical and non-medical members and a lay person, and its decisions on approval, continuing review and compensation are recorded and auditable. Before the 2019 Rules, ethics committee registration had been required under the old Rules since 2013, but the consolidation mattered, because the committee is the one actor in the trial system whose incentives are supposed to point at the participant alone.

Compensation is where the 2019 Rules made their hardest choice, and an honest account has to give both sides of it. The history runs through a period in which trial participants, and the families of participants who died, had no reliable route to compensation at all, a gap exposed by litigation and parliamentary scrutiny in the years after 2010. The Rules answer with a complete code. Where a trial subject is injured, the sponsor must provide free medical management for as long as the injury requires or until it is established that the injury is unrelated to the trial. Where the injury is trial-related, financial compensation follows, decided by the Central Licensing Authority on the recommendation of an expert committee, computed under the formula the Rules prescribe, which scales the payment by the age of the participant and the degree of disability, with a distinct computation where the subject has died. The quantum is not negotiated between sponsor and family; it is fixed by the regulator on the committee’s advice, and the sponsor’s obligation to pay does not depend on proof of negligence. Industry voices, including the Organisation of Pharmaceutical Producers of India in its submissions on the draft rules, objected that a no-fault compensation regime of this breadth, combined with the deemed timelines being illusory in practice, would push global trials to other countries and deny Indian patients early access to research. Patient advocates, including the health movements that had campaigned through the litigation years, answered that access purchased by uncompensated risk imposed on poor trial volunteers is not a benefit but a transfer. The Rules as made stand closer to the advocates’ position on compensation and closer to the industry’s position on timelines, which is why both sides continue to describe the settlement as incomplete.

One gateway through the approval system is persistently misdescribed and should be stated exactly. The Rules permit the Central Licensing Authority, in defined circumstances, to waive or abbreviate the local clinical trial requirement for a new drug already approved and marketed in specified countries with mature regulatory systems, where the conditions the Rules prescribe are answered: among them, that the drug’s behaviour in Indian patients can be predicted from the existing evidence, that no major safety concern specific to the Indian population is in view, and that the applicant undertakes the post-marketing surveillance the Authority directs. The waiver is a discretion exercised case by case on the file, not an entitlement carried by a foreign approval, and the standing committee’s 2012 findings were directed at exactly the slippage between the two: approvals in which the waiver logic was applied without the file showing its conditions. The gateway’s legitimacy depends on that distinction, and reform proposals in the field have generally sought to enumerate the specified countries and conditions more tightly rather than to close the route, because a system that retested every approved molecule in full would spend its scarce trial capacity duplicating evidence instead of filling the gaps that are genuinely India-specific, in pharmacogenetics, in disease burden and in patterns of use.

Approval is not the end of the Centre’s interest in a product. A new drug’s permission carries post-marketing obligations: the holder must report adverse reactions, submit periodic safety update reports, and conduct such post-marketing surveillance or phase IV study as the Authority directs, and the Authority may suspend or cancel the permission if the safety picture changes. The national collection point for the adverse-event half of that system is the Pharmacovigilance Programme of India, launched in 2010 with its national coordination centre at the Indian Pharmacopoeia Commission, which receives reports from hospitals, professionals and patients and feeds signals back to the regulator. The programme is the system’s memory: a medicine’s file does not close at approval, and the ban and restriction orders made under Section 26A in later years have repeatedly begun as signals in that memory rather than as findings in any inspection.

The approval gate has one documented breach, and it is structural rather than personal. A fixed-dose combination, two or more drugs in one dosage form, is by the Rules a new drug requiring central approval, because the interaction of the ingredients is itself a safety question. Yet for decades state licensing authorities granted manufacturing licences for combinations that had never been placed before the Centre, treating the combination as a mere assembly of already approved ingredients, and the market filled with several thousand combinations that existed by state licence alone. The Centre’s attempt in 2007 to direct the withdrawal of several hundred such combinations was challenged by manufacturers and stayed in litigation, leaving the products in the market and the question unresolved. The prohibition exercise of 2016 to 2018, traced in the enforcement sections of this article, was the second attempt to close the same gap by the ban route rather than the approval route. The episode is the clearest demonstration in the Act’s modern history that the split is not a drawing-office abstraction: a permission the Centre never gave could be manufactured into existence by a state licence, and the products travelled nationally on the strength of it.

Two boundaries of the approval gate complete the picture. The first is the border itself: import of a drug requires both a registration certificate for the product and the overseas manufacturing site and an import licence, both granted centrally, and port officers of the CDSCO sample consignments at entry against the import chapter’s definitions. The registration certificate is the distinctive instrument: it ties the Indian permission not merely to a product but to a named factory abroad, inspected or documentarily assessed, so that a change of manufacturing site is a regulatory event and not merely a commercial one. The bulk of India’s active pharmaceutical ingredients arrive through this gate from a concentrated set of overseas sources, a dependence that surfaced as a policy anxiety in its own right when supply chains tightened, and it is policed by the same registration and port-sampling machinery rather than by any separate statute. The second boundary is intellectual property, which the Act does not confer. Central approval is a safety and efficacy judgment, not a property right; it neither grants a monopoly in the molecule nor protects the applicant’s data from a later generic applicant who satisfies the Rules independently, a position long contested by originator companies and defended by the generic industry. The property questions live in the patent statute, and are treated in the account of medicine patents under the Patents Act; the regulatory gate and the patent gate are parallel, cumulative and frequently confused in public argument.

The import chapter at work: the Act at the border

Chapter III deserves a section of its own, because the border is the one point in the system where the divided machine does not operate: import control is wholly central, and it was the first part of the Act to commence. Section 10 lists what may not cross. A drug that is not of standard quality, a misbranded, adulterated or spurious drug under the import definitions of Sections 9 to 9D, a drug whose import the Rules prohibit, a patent or proprietary medicine whose label does not disclose its true formula or list of ingredients, and preparations whose containers carry exaggerated therapeutic claims, are all prohibited imports. Section 10A adds the reserve power: the Central Government may prohibit the import of any drug altogether where public interest requires it. Section 11 then borrows the customs machinery wholesale, treating a prohibited drug as prohibited goods under the law relating to sea customs, so that customs officers may detain and seize consignments and the confiscation provisions of the customs law apply; Section 14 completes the scheme by requiring the confiscation of a drug imported in contravention of the chapter.

The operative instruments are the registration certificate and the import licence, both granted by the Central Licensing Authority and both policed at a small number of port offices where CDSCO officers are posted alongside customs. The certificate registers the product and the named overseas site; the licence authorises the Indian importer to bring the registered product in. Consignments are liable to be sampled and sent for test, and the import definitions allow the same characterisation at the quay that the domestic definitions allow in the market. The chapter also makes room, under licence conditions administered through the Rules, for small quantities imported for personal use, for examination and test, and for government and charitable hospitals, the exceptions without which a border regime for medicines would obstruct the patients it exists to protect.

The chapter’s silence in the outward direction is as significant as its content in the inward one. The Act is a gatekeeper for the Indian market, not a regulator of Indian exports: it does not license the export of medicines as such, and the quality warranty an Indian exporter gives the world runs principally through the law and the regulator of the importing country, through the World Health Organization certification scheme, under which the central authority issues certificates of pharmaceutical product attesting that a product is licensed in India and its manufacturer inspected, and through the exporter’s own registrations abroad. The asymmetry explains a recurring diplomatic embarrassment of the system. When an Indian-made medicine harms a patient in another country, the failure is laid at the door of a statute that never claimed to police the transaction, while the instruments that did govern it, the state manufacturing licence, the Schedule M inspection and the importing country’s own gate, escape the headline. The pre-export testing direction for cough syrups from June 2023, imposed after the contamination deaths of 2022, was the system’s first general bridge across that gap, and it was built administratively, by direction under the existing licensing machinery, because the chapter itself contains no export gate to close.

Two features of the chapter repay attention. The first is its antiquity of design: it is recognisably the import code the Chopra Committee recommended, administered by an organisation the committee would also recognise, and its centralisation has never been controversial, because no state ever claimed the border. The second is the penalty survival noted in the definitions section: the chapter’s fines stand at their mid-century figures because no amendment has rebuilt them. The system’s practical answer has been administrative rather than penal: the consignment is stopped, the registration suspended, the goods confiscated or destroyed, and the importer who offends repeatedly finds the gate closed by licence action long before any court is troubled. Chapter III therefore exhibits, in miniature, the working constitution of the whole statute: formidable administrative control resting on a penal text of uneven vintage.

Licensing in the states: manufacture, loan licences and sale

Section 18 is the hinge between approval and the market. It prohibits the manufacture for sale or distribution, the sale, stocking, exhibition or distribution of any drug that is not of standard quality, or is misbranded, adulterated or spurious; of any drug manufactured in contravention of the Act or the Rules; and, in the clause on which the entire licensing system hangs, of any drug except under and in accordance with the conditions of a licence issued under the chapter, by Section 18(c). The licence is therefore not an administrative ornament. Manufacture or sale without one is itself an offence, ranked on the penalty ladder beside the quality offences, and the conditions written into a licence have the force of the Act behind them.

The manufacturing licence is granted by the licensing authority of the state in which the factory sits, on an application that must show premises conforming to the factory requirements of the Rules, competent technical staff with the qualifications the Rules prescribe supervising manufacture and testing, and an adequate testing facility or access to an approved laboratory. One refinement qualifies the picture of pure state control, and it matters. For specified sensitive categories, including blood and blood products, large-volume parenterals, vaccines and sera and recombinant DNA products, the licence is granted by the state authority only after approval by the Central Licence Approving Authority, the Drugs Controller General: a hybrid in which the Centre examines the capability and the state issues the paper. The categories chosen for this double key are exactly those where a manufacturing failure is least survivable by the patient, which is a tacit admission, inside the Rules themselves, that the undivided state licence is not thought sufficient for the highest-risk products.

The whole journey of a medicine, from molecule to counter, runs through both halves of the divided machine in sequence, and the steps are worth setting out in one place, because the gap between approval and licence is visible only when the two are laid side by side.

Step	Granted by	What the applicant must show	Where enforcement commonly fails
Clinical trial permission	Central Licensing Authority (Drugs Controller General)	Protocol, investigator and ethics committee particulars, preclinical and safety data	Trials approved on thin expert scrutiny; compensation disputes after injury
New drug approval	Central Licensing Authority, advised by Subject Expert Committees	Safety and efficacy evidence for the claims proposed; manufacturing and quality data	Post-marketing surveillance conditions weakly tracked after approval
Manufacturing licence	State licensing authority, with Central Licence Approving Authority concurrence for blood products, vaccines, sera, large-volume parenterals and recombinant DNA products	Premises and equipment meeting Schedule M, qualified technical staff, testing facilities	Factories licensed by hard-pressed state authorities and inspected rarely
Loan licence	State licensing authority	A lending unit with a valid manufacturing licence and spare capacity	Accountability split between brand owner and actual manufacturer
Sale licence, wholesale and retail	State licensing authority of each state of operation	Licensed premises, registered pharmacist supervision for retail sale, record systems	Identical product answering to dozens of authorities; records not maintained
Post-market sampling	Inspectors of the Centre and the states, testing by Government Analysts and the Central Drugs Laboratory	The product itself, drawn from the market, must meet the pharmacopoeial standard	Slow analysis and prosecution delay defeating the Section 25 retest right

What is a loan licence?

A loan licence lets a firm with no factory get its product made in another manufacturer’s licensed premises, using that unit’s equipment and technical staff. The licensing authority issues it where the lending unit holds a valid manufacturing licence and has spare capacity, and the loan licensee remains responsible for the quality of what is made under its name.

The loan licence is one of the most Indian institutions in the Act’s operation, and among the most criticised. It separates the brand from the factory: the company whose name is on the strip, which designed the marketing and takes the margin, need own nothing but a licence and a contract. The arrangement has a legitimate function, since it lets small firms and new entrants use spare capacity in established plants rather than duplicating capital, and the manufacturing unit’s licence conditions and inspections do attach to the premises where the product is actually made. Its pathology is accountability. When a batch made under a loan licence fails, the inspectorate must chase two firms in possibly two jurisdictions, and each can present the other as the party in fact responsible for the defect. Official reviews have repeatedly identified third-party and loan-licensed manufacture as a point of systemic weakness, and the tightening of conditions for such licences has been a standing item of regulatory reform.

Sale licensing completes the chain. A wholesale licence is required to sell, stock or distribute drugs to retailers and institutions; a retail licence is required to sell to the patient, and retail sale of the prescription categories must be under the supervision of a registered pharmacist, against the prescription the schedules demand. Licences are premises-specific and jurisdiction-specific, and renewal, retention and the register of licences are all administered by the state authority. The licence lifecycle changed character in 2017, when an amendment to the Rules replaced the fixed-term licence, expiring and requiring renewal by fresh application, with a licence retained in perpetuity on payment of a retention fee every five years, unless suspended or cancelled in the meantime. The change was an ease-of-business measure, and its regulatory significance is easily missed: the periodic renewal application had been the one moment at which the licensing authority routinely re-examined every licensee on its register, and its abolition transferred the entire weight of continuing supervision to inspection. A system of perpetual licences policed by a thin inspectorate drifts towards a register of the unexamined, and the risk-based inspection programmes described later, in which premises are graded and the weakest inspected first and jointly, are best understood as the system’s attempt to rebuild, in targeted form, the checkpoint that renewal once supplied universally.

Licensing has also acquired a digital layer that the Act never mentions and the Rules only partly describe. Central permissions, import licences, registration certificates and trial permissions are applied for and issued through the CDSCO’s online portal, and state authorities have progressively moved applications, licence records and inspection reporting onto their own portals and onto national dashboards that let the Centre see, for the first time, a consolidated picture of the licences its code has generated. The change is more than clerical. A paper licence system made the state register the only record and the physical file the only audit trail; a portal system creates data, and data is what risk-based inspection, the recall of a failed batch across state lines, and the detection of a manufacturer licensed in several states under variant names, all feed on. The statute’s text still imagines an inspector with a seizure memo; the system he works in increasingly begins from a database query, and the gap between those two pictures is the distance the unrepealed Act has travelled from its own drafting.

Why does one medicine need licences from more than one state?

Because a sale licence speaks only for the territory of the authority that granted it. A distributor stocking the same strip in three states answers to three state licensing authorities, holds licences under each, and faces inspection by each, although the product, the manufacturer and the central approval behind it are identical in all three places.

The disciplinary side of licensing runs on administrative rather than criminal rails, and for most licensees it is the side that matters. The licensing authority that finds a contravention may, after giving the licensee an opportunity to show cause, suspend or cancel the licence for the period it thinks fit, and the Rules give the aggrieved licensee an appeal to the state government. Suspension of a manufacturing licence stops the factory as effectively as any conviction and far faster; it is the sanction the industry fears most and the one the official statistics record most often, in the form of stop-production orders, suspensions and cancellations issued after inspections. The history of the system’s enforcement is therefore double: a criminal ladder of genuine severity, climbed rarely, and an administrative ladder of licence action, climbed constantly. Any account that reads only the penalties in Section 27 will misunderstand how the Act actually disciplines its licensees.

Manufacturing quality is policed through Schedule M, the good manufacturing practices schedule to the Rules, which prescribes the premises, plant, equipment, documentation, quality-control and personnel requirements for each dosage form, and against which factories are inspected. The schedule was comprehensively rewritten by notification of 28 December 2023, which replaced the older good manufacturing practices text with an expanded code built around a pharmaceutical quality system, quality risk management and annual product quality review. The notification staggered its own commencement by the size of the firm: manufacturers with annual turnover above 250 crore rupees were given six months from publication, and smaller manufacturers twelve months, a timetable later softened for small and medium manufacturers that filed upgradation plans, with the extended date running to the end of 2025. The revision was the Centre’s direct answer to the export quality failures discussed in the enforcement section, and its staggered design concedes the federal objection in advance: a single national standard, imposed on thousands of small factories by a fixed date, would close units faster than it would improve them.

The Act at the counter: retail practice and the traceable chain

The retail chemist is the point at which the entire apparatus becomes visible to the citizen, and the Act regulates that counter with the same licensing logic it applies to the factory, plus one device of particular elegance: the compulsory paper trail. A retail licence is granted for named premises, and its conditions require that the sale of medicines be conducted under the personal supervision of a registered pharmacist, that prescription medicines be supplied only against the prescriptions the schedules demand, that records of purchases and sales be preserved, and that the premises and storage meet the conditions the Rules prescribe. The pharmacist’s presence is not decorative. The Pharmacy Act 1948 reserves dispensing to the registered profession, and this Act borrows that reservation as a licence condition, so that the last transaction in the chain is legally a professional act and not merely a sale.

The paper trail is Section 18A. Every person who sells or distributes a drug must disclose to an Inspector, on demand, the name, address and particulars of the person from whom he acquired it, and Section 28A backs the record-keeping duties with a penalty for failure to keep documents or furnish information. Working together with the purchase invoices a licensee must preserve, these duties convert every strip on every shelf into the end of a traceable chain: inspector to retailer, retailer to wholesaler, wholesaler to manufacturer, each link identified by a document the holder was legally obliged to keep. When a spurious strip is found in a patient’s hands, the investigation is, in form, nothing more than walking that chain backwards until it reaches the link that cannot produce its paper, which is usually the link where the counterfeit entered. The chain is also why the record offences, trivial as they read beside the spurious ladder, are prosecuted seriously: a break in the paper is frequently the only provable offence a well-run counterfeiting operation commits in the open.

Homoeopathic medicines occupy a fourth stream beside allopathic drugs, Ayurvedic products and cosmetics, governed through dedicated provisions of the Rules under which their manufacture and sale are licensed, a reminder that the Act’s perimeter has always been drawn by enumeration and notification rather than by any single theory of what a medicine is. The streams share the inspectorate, the analysts and the penalty philosophy, which is why a chemist’s shop may hold, on adjoining shelves, products answerable to four different regulatory logics under one roof and one licence-holder’s signature.

The Ayurvedic code within the code

Chapter IVA, inserted in 1964, is a statute inside the statute, and a complete account of the Act must give it its own paragraph of architecture. Ayurvedic, Siddha and Unani medicines were brought under control late and on different terms, because the quality question they pose is different: their standards are fixed not by a pharmacopoeia of assay values but by the authoritative texts of the systems themselves, the First Schedule to the Act, which lists the classical works by reference to which a formulation’s identity is judged. The chapter carries its own definitions of misbranded, adulterated and spurious ASU drugs in Sections 33E, 33EE and 33EEA, its own prohibition section, its own inspectors and Government Analysts notified for the purpose, and its own licensing authority structure, under which manufacture is licensed by the state authority and good manufacturing practice is prescribed by Schedule T to the Rules. Penalties under the chapter, in Section 33I, are markedly lighter than the Section 27 ladder, a difference its critics read as a hierarchy of seriousness between medical systems and its defenders read as proportionate to formulations whose risks are ordinarily of adulteration and contamination rather than of pharmacological potency. The friction point of the modern chapter is the patent or proprietary ASU medicine, a branded formulation that is not a classical text preparation, which must pass through a licensing and, in the case of new claims, an evidence discipline that critics describe as steadily converging on the allopathic model. The chapter also operates alongside the advertisement controls carried in the Rules and the separate statute of 1954 on objectionable advertisements, which together police the cure claims with which such products have historically been marketed. A reader who masters Chapter IV alone knows perhaps three quarters of the Act; the remaining quarter is this parallel code, administered by much of the same machinery in the name of a different theory of what a standard is.

The schedules, where practice actually lives

If the Act is the skeleton, the schedules to the Rules are the working organs, and a pharmacist or regulatory manager thinks in their letters the way a litigator thinks in sections. Schedule C and Schedule C1 list the biological and special products, sera, vaccines, insulin and the like, whose manufacture and testing attract special provisions, including batch testing. Schedule F prescribes standards for particular products such as vaccines and surgical dressings. Schedule G lists medicines that must carry the caution that they are dangerous to take except under medical supervision. Schedule J, read with the rule that gives it effect, names diseases and conditions for which no drug may claim to prevent or cure, the statutory wall against miracle-cure advertising. Schedule P fixes the life period, the shelf life, of listed drugs. Schedule U prescribes the records of manufacture and sale that licensees must keep. The schedules that dominate daily practice, though, are the three prescription schedules.

Schedule H is the general prescription list: the medicines in it may be sold by retail only on the prescription of a registered medical practitioner, a status whose register and discipline are governed after 2019 by the National Medical Commission Act, and the pack must carry the prescription warning. The list runs to several hundred substances, and its presence on a strip is the legal difference between a medicine the chemist may hand over on request and one he may not. Schedule X is the sterner sibling, reserved for habit-forming and dependence-producing psychotropic substances: it demands a prescription in duplicate, special licences and records preserved for two years, and it carries the system’s anxiety about diversion into abuse.

Schedule H1 is the youngest of the three and the most instructive, because it shows the schedule mechanism being used to answer a policy emergency. Inserted by notification G.S.R. 588(E) of 30 August 2013 and in force from 1 March 2014, it began with forty-six named substances: principally higher-generation antibiotics such as the later cephalosporins and fluoroquinolones, the anti-tuberculosis drugs including rifampicin and isoniazid, and habit-forming medicines such as codeine, tramadol and zolpidem. The target was antimicrobial resistance and the casual over-the-counter sale of antibiotics, a practice prescription law on paper had failed to stop. Schedule H1 therefore adds the audit trail Schedule H lacks: the chemist must record each supply in a separate register showing the name and address of the prescriber, the name of the patient, the drug and the quantity, and must keep the record for three years open to inspection, while the pack carries a red-bordered warning box and the red Rx symbol. A prescription schedule with a memory is a different instrument from a prescription schedule without one, and the inspectors’ ability to reconstruct a sale three years later is precisely what the register buys.

Clinical research had its schedule too. Schedule Y governed the conduct of trials for a generation, and its displacement by the New Drugs and Clinical Trials Rules 2019, described in the approval section of this article, is the standing proof of the series thesis: the most ethically charged chapter of drug law was carried for thirty years by a schedule, and when it finally outgrew that form it moved not into the Act but into a new set of rules. The Act itself was not amended to receive it. Nothing in the pattern has changed since: the statute stands still while the law moves underneath it by notification.

One schedule discipline cuts across all the others and deserves separate mention, because it is the closest the system comes to testing a product before rather than after it reaches patients. For specified biological products, vaccines, sera and blood products among them, the Rules subject manufacture to batch discipline: samples of batches are liable to be drawn and tested through the central laboratory network, and the Central Drugs Laboratory and the institutes notified for the purpose conduct the lot release testing on which market release of the most sensitive products depends. The ordinary drug reaches the shelf on the strength of its licence and is tested afterwards, in the market, by sampling; the scheduled biological is tested as a batch, in advance, as a condition of its ordinary circulation. The distinction maps the system’s own judgment about where after-the-fact enforcement is tolerable and where it is not: a substandard tablet is a prosecution, while a contaminated vaccine batch is a catastrophe, and the schedules quietly encode that difference in kind.

Medical devices inside a drugs statute

The device sector arrived under this Act by the side door, and it has never stopped arguing that the door is the wrong shape. Section 3(b)(iv) admits, as drugs, only such devices as the Central Government specifies by notification after consulting the Board. For decades the specification was miserly: from the mid-2000s a small set of sterile devices and implants was notified, and by the end of the following decade only twenty-three categories of devices stood regulated at all, in a market selling everything from syringes to imaging systems. The Medical Devices Rules 2017 created a purpose-built regime within the Act’s shelter, classifying devices by risk into Classes A to D and prescribing licensing proportionate to class, but they applied only to the notified categories.

The notification of 11 February 2020, S.O. 648(E), changed the perimeter at a stroke: with effect from 1 April 2020, all devices, instruments, apparatus, appliances, implants, materials, software and accessories intended for the specified medical purposes, and not acting primarily by pharmacological, immunological or metabolic means, were specified as drugs. Registration and then licensing were phased in by risk class over the following years. The regime the Rules built for devices is risk-proportionate in a way the drug regime is not, and the contrast is instructive. Class A covers the lowest-risk devices, tongue depressors and examination gloves; Class B the low-to-moderate range, including most diagnostic kits; Class C the moderate-to-high range, including ventilators and orthopaedic implants; and Class D the highest, including heart valves and implantable defibrillators. Licensing follows the class: the lower classes are licensed by the state authorities, the higher classes by the Centre, and manufacture must rest on a documented quality management system rather than on the pharmaceutical good manufacturing practices of Schedule M, because the discipline that makes a sterile implant safe is a design and process discipline, not an assay discipline. A dedicated chapter of the system therefore already exists inside the Act’s shelter, with its own regulator split, its own conformity logic and its own audit culture, and it functions as a standing demonstration that the parent statute can host a differently shaped regime when the Rules are written for the object being regulated rather than borrowed from the object next door.

The consequences run both ways, as the sector’s critics and defenders each observe. A syringe and a medicine are not the same kind of object: device failure is usually an engineering and quality-system failure, not a chemical one, and the criminal machinery of a drugs statute, designed for poisons and counterfeits, sits awkwardly on a software update or a design defect. The Association of Indian Medical Device Industry has argued the point for years, and a parliamentary committee recommendation for a separate medical devices law stands on the record. The counter-argument is actuarial: the notified route delivered a functioning, if ill-fitting, regulator immediately, and a bespoke statute would have taken years the market did not have, as the still-unenacted replacement bill demonstrates. Devices are therefore regulated as drugs not because anyone thinks they are drugs, but because the notification power in a 1940 definition made it possible in a single gazette issue, and no Parliament has yet made anything better.

Enforcement: inspectors, samples and the penalty ladder

The enforcement design is prosecutorial and sample-driven. Its central ritual is the drawn sample, and the rights and duties around that ritual decide more cases than the definitions do. An Inspector appointed under Section 21 may enter and search any premises where a drug or cosmetic is manufactured, sold or stocked, inspect the plant and the records, and, under Section 23, take samples of any product for test or analysis. The sampling procedure is prescribed in forensic detail: the Inspector pays or tenders the price, divides the sample into four portions, seals and marks each in the presence of the person from whom it is taken, restores one portion to that person, sends one to the Government Analyst, and retains the remainder so that one portion can be produced in court and one can go, if the dispute ripens, to the Central Drugs Laboratory.

What happens to a drug sample after an inspector draws it?

The Government Analyst tests the portion sent and reports to the Inspector. The report is evidence, but not the last word: the person charged may, within twenty-eight days, notify the court of an intention to controvert it, and the court sends the reserved portion to the Central Drugs Laboratory, whose Director’s report is conclusive evidence of the facts it states.

Section 25 is the accused’s charter and the prosecution’s trap, and it rewards close statement. The Government Analyst’s report is furnished to the Inspector and a copy goes to the person from whom the sample was taken or to the person whose name and address the seller disclosed under Section 18A. Unless that person, within twenty-eight days of receiving it, notifies the Inspector or the court in writing of an intention to adduce evidence in controversion of the report, the report stands as conclusive evidence of the facts stated in it. If the intention is notified, Section 25(4) obliges the court to have the reserved sample portion tested by the Central Drugs Laboratory, and that report is conclusive. The whole safeguard decays with time, because medicines expire: a sample sent for retesting after its shelf life has run cannot be tested at all, and the conclusive first report hardens into the case. The Supreme Court enforced exactly that logic in Medicamen Biotech Limited v. Rubina Bose, Drug Inspector, decided in 2008, where the complaint had been filed only about a month before the drug’s expiry and the machinery moved so slowly that retesting became impossible; the Court held that the manufacturer had been deprived of the valuable right conferred by Sections 25(3) and 25(4), and quashed the proceedings. The case is the standing warning to every drugs inspector in India that delay is not a neutral administrative fact in this statute but a ground of acquittal.

Prosecutions under Chapter IV are gated by Section 32. No prosecution may be instituted except by an Inspector, by any gazetted officer of the Central Government or a State Government authorised in writing for the purpose by general or special order, by the person aggrieved, or by a recognised consumer association. Offences under the chapter are tried by no court inferior to a Court of Session since the amendment of 2008, and the same amendment gave the Central and state governments power, under Section 36AB, to designate Courts of Session as Special Courts for the offences relating to adulterated and spurious drugs, and inserted Section 36AC, which makes the listed offences relating to adulterated and spurious drugs cognizable, and bars bail unless the public prosecutor has been heard and, where bail is opposed, the court is satisfied there are reasonable grounds for believing the accused is not guilty and is unlikely to offend while on bail. The bail clause deliberately echoes the stringent bail architecture of the special criminal statutes: for the counterfeit medicine trade, Parliament chose to reverse the ordinary presumptions.

The penalties for the Chapter IV quality offences sit in Section 27, and their ladder, as recast by the Drugs and Cosmetics (Amendment) Act 2008, Act 26 of 2008, brought into force on 10 August 2009, runs as follows. At the top, clause (a): whoever manufactures for sale, sells, stocks or distributes an adulterated or spurious drug which, when used, is likely to cause death or grievous hurt, solely on account of being adulterated or spurious or not of standard quality, is punishable with imprisonment for not less than ten years, extendable to imprisonment for life, and a fine of not less than ten lakh rupees or three times the value of the drugs confiscated, whichever is more. Two provisos then do something unusual in Indian penal drafting: the fine realised is directed to be paid by way of compensation to the person harmed, and where the user has died, to the relative of the deceased, a word the section itself defines. Clause (b) covers the adulterated drug that does not fall within clause (a), and manufacture or sale without the licence Section 18(c) requires: imprisonment of not less than three years and up to five years, with fine of not less than one lakh rupees or three times the value of the confiscated drugs. Clause (c) covers the spurious drug outside clause (a): imprisonment of not less than seven years, extendable to life, with fine of not less than three lakh rupees or three times the confiscated value. Clause (d) is the residual rung for other contraventions of the chapter: imprisonment of one to two years and fine of not less than twenty thousand rupees. Courts may go below the stated minima only for adequate and special reasons recorded in the judgment. Section 27A carries a parallel ladder for cosmetics, Section 28 punishes failure to disclose the name and address of the manufacturer from whom goods were obtained, Section 28A the failure to keep documents or furnish information, and Section 30 raises the exposure of the repeat offender. Section 34 addresses offences by companies, attributing liability to every person who was in charge of, and responsible to, the company for the conduct of its business when the offence was committed, together with the company itself, and reaching directors, managers and officers where the offence occurred with their consent, connivance or neglect.

Against the ladder stand the defences, and Section 19 contains the one that decides the fate of the honest trader. A seller who is prosecuted for a drug that proves adulterated, spurious or not of standard quality may defend the charge by proving that the product was acquired from a duly licensed manufacturer, distributor or dealer, under a written warranty of quality of the kind Section 19 prescribes; that there was no reason to believe the product contravened the Act; and that it was sold in the same state in which it was received, properly stored throughout. The warranty defence is the Act’s allocation of responsibility along its own chain: the retailer who bought in good faith from a licensed source, kept the paper and stored the goods properly is not the criminal the chapter is hunting, and the prosecution is redirected to the warranty-giver up the chain. Its conditions explain the behaviour of every careful chemist in India, who will not buy from an unlicensed source, will not buy without the invoice and warranty the defence requires, and treats a missing document as a reason to refuse stock. A companion provision protects the employee or agent who acted without knowledge and with due diligence, leaving the liability with the principal who directed the business.

The Act thus runs two enforcement tracks against the same conduct, and their interplay decides most real disputes. The criminal track, prosecution under Section 27 on an analyst’s report, is slow, proof-heavy and reserved in practice for the counterfeit and the dangerous. The administrative track is faster and, for a manufacturer, often the more feared: the licensing authority may, after giving the licensee an opportunity to show cause, suspend or cancel the licence for a stated period or altogether, may direct production to stop, and may order the recall or destruction of offending stock, and these orders take effect without waiting for any prosecution to conclude. The tracks are independent; an acquittal does not restore a cancelled licence, and a compounded offence does not conclude the licensing question. The natural justice discipline on the administrative track, show cause, a hearing in substance, a reasoned order, and appeal to the government the Rules designate, is judge-made and rule-made rather than spelled out in the Act, and the writ jurisdiction of the High Courts is where its boundaries are litigated. For the regulated firm the practical statute is therefore double: a criminal code of great severity that moves slowly, and an administrative code of moderate severity that can close the factory first.

Section 34, the offences-by-companies provision, determines who stands in the dock beside the firm, and its structure repays statement because it is misreported as often as any provision in the Act. Where an offence is committed by a company, every person who at the time was in charge of and responsible to the company for the conduct of its business is deemed guilty along with the company itself, subject to the proviso that protects the person who proves the offence was committed without his knowledge or that he exercised all due diligence to prevent it. A second limb reaches directors, managers, secretaries and officers where the offence occurred with their consent or connivance or is attributable to their neglect. The Supreme Court’s discipline in Brij Lal Mittal, noticed earlier, works on the first limb: designation is not responsibility, and the complaint must plead the factual basis on which the accused officer was in charge of the business. The provision interacts with the warranty defence and the traceable chain to give the Act its characteristic liability geometry: responsibility settles on the person who controlled the transaction, the manufacturer for the formulation, the licence-holder for the premises, the officer in charge for the company, and it slides off the intermediary who bought in good faith, kept the paper and can point up the chain to the warrantor.

Two qualifications keep the ladder honest. The first is that its most severe rungs are climbed slowly and seldom: prosecutions must run the sampling gauntlet, survive the Section 25 retest right, and be proved category by category, and the disposal of drug cases in the courts is measured in years. Severity on paper and certainty in practice pull in opposite directions, and the system’s real deterrent is distributed unevenly between the exemplary prosecution and the routine licence suspension. The second qualification is administrative practice. Guidance issued through the CDSCO for action on failed samples divides findings into categories: the spurious and adulterated at the top, where prosecution and the Section 36AC machinery are the default; grossly substandard products from licensed manufacturers in the middle, where negligence and good manufacturing practice failures are examined; and minor defects at the bottom, broken or discoloured tablets, labelling errors and variations traceable to formulation or storage, where suspension, cancellation or compounding is ordinarily preferred to a prosecution that would not serve the ends of justice. The guidance is the system’s own admission that a single ladder cannot justly price both the counterfeiter and the careless.

Compounding, the statutory settlement of minor offences by payment, arrived with the 2008 amendment as Section 32B, and was significantly widened by the Jan Vishwas (Amendment of Provisions) Act 2023, whose changes to this Act took effect from 31 December 2024. The Jan Vishwas amendments to the 1940 Act were narrow and should be stated exactly, because they are often described as a general decriminalisation of drug offences, which they are not. They touched three provisions only. Section 29, which punishes the use of a Government Analyst’s report in advertising a drug or cosmetic, had its fine converted to a penalty of one lakh rupees. Section 30, dealing with repetition of that advertising offence, was put on a purely monetary footing with a penalty extending to five lakh rupees, imprisonment being removed for that offence. And Section 32B was amended to add the residual-rung offences under Section 27(d) and the corresponding cosmetics clause, Section 27A(ii), to the list of offences that may be compounded, alongside the import, disclosure and record offences already there. Nothing in the Jan Vishwas Act touches the imprisonment prescribed for adulterated or spurious drugs, and offences once compounded cannot be compounded again on a second conviction. The compounding machinery was then given procedural form by the Drugs and Cosmetics (Compounding of Offences) Rules finalised in April 2025.

The compounding route deserves its mechanics, because it is where the decriminalisation policy actually happens. An accused person or firm applies to the compounding authority designated under the Rules, before the prosecution is instituted or, with the court’s leave, after it has begun. The authority hears the applicant, fixes a compounding amount within the framework the Rules prescribe, and on payment the applicant is discharged and gains immunity from prosecution for the compounded offence. The immunity is conditional in substance as well as in form: the Rules empower the authority to withdraw it where the applicant conceals material particulars, gives false evidence or fails to cooperate, and the amount once paid is not refundable merely because the applicant later regrets the bargain, though it is returned where the court declines to accept the immunity. The design is borrowed from the tax and corporate compounding regimes and its logic is prosecutorial triage: the inspectorate’s scarce court time is reserved for the adulterated and spurious ladder, where compounding is not available at all, while the record offences, the advertising offences and the residual rung are cleared for money. Whether the exchange is wise depends on how often the minor offences are, in truth, the visible edge of a manufacturing failure, a question the published record does not yet answer.

Can the Centre ban a medicine that is already being sold?

Yes. Section 26A empowers the Central Government, where it is satisfied that a drug involves risk to human beings or animals, or lacks the therapeutic value claimed for it, or contains ingredients in a quantity for which no therapeutic justification exists, to prohibit its manufacture, sale and distribution in the public interest by gazette notification.

The ban power has its own litigation history, which tests the deference courts owe to expert judgment in this field. In March 2016 the Centre prohibited 344 fixed-dose combinations, acting on the report of an expert committee chaired by Professor Chandrakant Kokate, and the Delhi High Court struck the prohibitions down in December 2016 for want of consultation with the Drugs Technical Advisory Board and for the manner of their adoption. The Supreme Court, in December 2017 in Union of India v. Pfizer Limited, declined to restore the bans outright and instead directed that the combinations be examined by the Board, and on the Board’s recommendations the Centre in September 2018 prohibited 328 of the combinations. The sequence teaches the operating rule of Section 26A precisely: the power is wide, and the Supreme Court held that prior consultation with the Board is not a statutory condition of its exercise, but the government’s satisfaction must answer the section’s own tests, risk to human beings or animals, absence of the therapeutic value claimed, or ingredients in quantities for which there is no therapeutic justification, and a prohibition resting on no material directed to those tests will not stand. The older authority runs the same way. In Vincent Panikurlangara v. Union of India, decided in 1987, the Supreme Court, dealing with a public interest petition against harmful and irrational formulations, treated the identification of drugs fit to be withdrawn as a task for expert bodies advising the government, and measured the Centre’s action by that standard. And on the company-liability side, the Supreme Court’s reading of Section 34 in State of Haryana v. Brij Lal Mittal in 1998 insists that a director or officer is not liable by designation alone: the prosecution must plead and prove that the person was in charge of and responsible for the conduct of the business, a discipline that has structured charge-sheets under the Act ever since.

Taken together, the case law teaches a consistent lesson about where this statute’s strength and weakness lie. Courts have not diluted the penalty ladder; no line of authority reads down the Section 27 rungs or softens the bail discipline for the counterfeiting offences, and the deference line from Panikurlangara to Pfizer protects expert-led prohibition decisions taken on proper material. What the courts police, with equal consistency, is the process by which the state builds its case and takes its decisions: the sampling formalities of Section 23, the twenty-eight day controversion window and retest right of Section 25, the in-charge requirement of Section 34, the consultation precondition where the statute imposes one, and the requirement that even the wide Section 26A satisfaction rest on material answering the section. The pattern is the opposite of the lay assumption that drug cases fail because judges are lenient. They fail, when they fail, because a statute that makes paperwork the spine of proof will acquit on broken paperwork, and the reported decisions read collectively as an instruction manual addressed to inspectors, analysts and licensing officers rather than to the accused.

What the Act does not do

A pillar article owes its reader the statute’s boundaries as clearly as its contents, and two boundaries are chronically misstated. The Act does not license medical practice, and it does not control price. A medicine’s legality under this statute and its affordability are governed by different instruments held by different hands, and the confusion between them produces a large share of the misdirected anger in Indian drug debates.

Does the Drugs and Cosmetics Act control the price of a medicine?

No. The Act controls quality, safety and the conditions of manufacture and sale; it contains no price-fixing power. Medicine prices are controlled under the Drugs (Prices Control) Order, made under the Essential Commodities Act, and administered by the National Pharmaceutical Pricing Authority, a separate regime with a separate logic.

Price control reaches medicines because the Central Government, acting under Section 3 of the Essential Commodities Act 1955, has declared drugs an essential commodity and made the Drugs (Prices Control) Order, the 2013 Order being the operative version, under which ceiling prices are fixed for the formulations in the National List of Essential Medicines and the prices of other formulations are monitored against steep annual increases. The wider architecture of that essential commodities power, and the way an order made under it borrows the machinery of another field entirely, is the subject of the account of the Essential Commodities Act; here the point is jurisdictional and simple. When a ceiling price is breached, the overcharging is recovered under the prices Order, not prosecuted under Section 27, and when a medicine is spurious, the prices Order has nothing to say. The same tablet can be lawful under both regimes, unlawful under either, or unlawful under both, and a reader who cannot separate the two questions will misread every Indian drug controversy, most of which are about one regime reported in the vocabulary of the other.

The prices regime has its own archaeology, worth a paragraph because its instruments are so often attributed to this Act. Price control over drugs began with the first Drugs (Prices Control) Order of 1970, and the modern framework is the 2013 Order, under which the National Pharmaceutical Pricing Authority, constituted in 1997 under the Department of Pharmaceuticals, fixes ceiling prices for scheduled formulations, those in the National List of Essential Medicines, by a market-based formula averaging the prices of brands with significant share, and monitors the prices of non-scheduled formulations against a ceiling on annual increase. The 2013 Order also carries, in its paragraph 19, an extraordinary power to fix prices in the public interest outside the schedule, exercised most prominently to cap the prices of coronary stents in 2017 and of orthopaedic knee implants later that year. Each element of that apparatus, the list, the formula, the extraordinary power, belongs to the essential commodities line of legislation. Nothing in it amends the 1940 Act, and a medicine fully compliant with the prices Order may still be spurious under Section 17B, which is the whole of the point about boundaries.

The split indicted: Mashelkar, the standing committee and the centralisation debate

The structural criticism of the Act is not a journalistic trope; it is the considered finding of the two most authoritative examinations of the system, and any defence of the statute has to answer them. The first was the Expert Committee on a Comprehensive Examination of Drug Regulatory Issues, Including the Problem of Spurious Drugs, constituted by the Ministry of Health and Family Welfare on 27 January 2003 under the chairmanship of Dr R. A. Mashelkar, then Director General of the Council of Scientific and Industrial Research. Its report, submitted in November 2003, located the system’s failures not in the statutory text but in the regulatory infrastructure beneath it: inadequate and weak drug control organisations at the Centre and in the states, too few inspectors, too few testing laboratories, and the absence of a strong central authority. Its central institutional recommendation was the conversion of the CDSCO into an empowered, professionally managed Central Drug Administration reporting to the Health Ministry. Its penal recommendations, enhanced punishment for spurious drugs, special courts, cognizable and non-bailable offences, compounding of minor offences, supplied the blueprint that the 2008 amendment enacted five years later. But the committee’s deepest point was administrative rather than penal: a statute enforced by an inspectorate that does not exist on the ground will fail at whatever penalty level Parliament sets.

The second examination was parliamentary. The Fifty-ninth Report of the Parliamentary Standing Committee on Health and Family Welfare, on the functioning of the CDSCO, tabled in May 2012, subjected the approval side of the split to a scrutiny the Mashelkar Committee had not attempted. The committee found that thirty-three new drugs had been approved for sale in India between January 2008 and October 2010 without clinical trials in India, and its examination of a sample of thirty-nine approvals found phase III trials missing where the Rules required them, expert opinion unrecorded or thin, and drugs approved that had no permission for sale in any of the major developed jurisdictions and no special relevance to Indian medical needs. It recorded, in language without parliamentary precedent in this field, its conclusion that the record disclosed a collusive nexus between drug manufacturers, some functionaries of the regulator and some medical experts. The report’s staffing findings were equally bleak: it found only 127 of the CDSCO’s 327 sanctioned posts filled, with a small cadre of deputy and assistant drugs controllers handling an annual workload the committee counted in tens of thousands of applications, alongside oversight of thousands of manufacturing units and hundreds of thousands of sales outlets. Between them the two examinations frame the system’s twin deficits: the Centre approves with too little capacity for the responsibility, and the states license and inspect with too little capacity for theirs.

The state’s numbers, when the system finally counted them honestly, were less catastrophic than the rhetoric and worse than the official comfort of earlier decades. The national survey of the extent of the problem of spurious and not of standard quality drugs, carried out for the Health Ministry by the National Institute of Biologicals over 2014 to 2016 and released in February 2017, tested 47,012 samples drawn from supply chains across the country and estimated 3.16 per cent of drugs to be not of standard quality and 0.0245 per cent to be spurious. The survey’s importance is methodological as much as numerical: for the first time the state possessed a defensible measurement, and the measurement showed a problem of widespread ordinary quality failure, the NSQ tail, rather than the flood of murderous counterfeits of popular description. Policy that treats the two as one problem will misfire at both, and the survey is the strongest evidence for the administrative, category-sensitive enforcement guidance described earlier in this article.

The export failures of 2022 tested the system from the other side. In that year, cough syrups manufactured in India were linked by the World Health Organization and by investigators in the importing countries to the deaths of children in The Gambia and Uzbekistan, the syrups being contaminated with diethylene glycol and ethylene glycol, industrial solvents whose presence is the paradigm of the adulterated category. The regulatory response ran along the Act’s existing rails rather than any new law: risk-based joint inspections of manufacturing premises by central and state officers, licence suspensions and stop-production orders against identified units, and, from June 2023, a requirement that cough syrups be tested at specified government laboratories before export. The episode is the dual structure in action and under indictment at once: the licences suspended were state licences, the standards breached were central standards, and the capacity that failed was the inspection capacity of the state where the syrups were made. The joint-inspection instrument has since been generalised: the Health Minister informed Parliament in 2024 that risk-based joint inspections by central and state teams had covered more than 400 manufacturing premises and produced over 300 regulatory actions, including licence suspensions and cancellations, stop-production orders and warning letters. The figures are a ministerial statement, to be weighed as such, but the direction they record is the one both expert examinations prescribed: the two halves of the machine inspecting together because neither half’s inspection, alone, had proved sufficient.

The debate about what to do with the split has two serious sides, and the stance this article owes the reader is to give each its strongest form before judging. The industry’s case, pressed by the Indian Pharmaceutical Alliance and the manufacturers’ associations in consultation after consultation, is that dual regulation manufactures duplication and delay: a firm answers to the Centre for its product and to a state for its factory, times its launches to divergent state practices, and watches identical questions answered differently across jurisdictions, while the licensing of manufacture by the weakest state authority sets the effective national standard for every product that state ships across the country. Forum shopping is not an allegation but an arithmetic consequence: a manufacturer refused in a strict state may apply in a lax one, and the product then travels everywhere under the commerce of a single market. The federal and capacity answer, pressed by state drug controllers and by the public-health movements, is that licensing is not a paper function but the legal shadow of inspection, and inspection happens in a place, by officers who can reach the factory. A central authority in Delhi cannot inspect tens of thousands of manufacturing units and hundreds of thousands of sales outlets; centralising the licence without multiplying the inspectorate would produce nominal uniformity and actual absence, a single queue in place of thirty-six uneven ones, and the states’ local knowledge of their own trade, accumulated in their enforcement cadres, would be discarded at the moment it is most needed.

The verdict the record supports is narrower than either side’s programme. The evidence of the two examinations and of the survey does not show that state licensing of sale is the system’s weakness; retail inspection is irreducibly local, and no serious proposal gains anything by moving it. The weakness concentrates at one node, the licensing of manufacture, because a factory’s output crosses every border the sale licences police. At that node the Act already contains the germ of the answer in the Central Licence Approving Authority device, which pairs central technical approval with state issue for the highest-risk categories, and in the joint inspection programmes that pair central and state officers in the same factory. Extending the double key to all manufacture, so that no factory ships nationally on the unexamined word of a single hard-pressed state authority, answers the forum-shopping objection without pretending that Delhi can inspect India. It is, in substance, the Mashelkar direction travelled by the road the Rules have already built, and it requires no new statute, only the notification habit this field already lives by.

The replacement that has not arrived

Every part of the machine described in this article operates under a suspended sentence, because the statute’s replacement has been drafted, published, debated and withheld for a decade. The first attempt was legislative in form: a Drugs and Cosmetics (Amendment) Bill introduced in 2013 proposed, among other things, to put clinical trials and medical devices on a statutory footing, and in 2016 the government withdrew it, the then Drugs Controller General explaining the decision in terms that have become the epitaph of the whole enterprise, that it was better to write a new law than to keep putting bandages on the old one. The second attempt was a clean-slate draft. In July 2022 the Ministry of Health and Family Welfare published the draft Drugs, Medical Devices and Cosmetics Bill 2022 for public comment: a consolidated statute that would regulate drugs, medical devices, cosmetics and Ayurvedic products within one text, govern clinical trials expressly rather than by notification, constitute a statutory central regulator, regulate online pharmacies for the first time, and extend the device sector’s coverage by enactment rather than by the notification route of 2020. A version of the bill was listed for introduction in Parliament in the monsoon session of 2023, and was not introduced. The draft remains a draft. It has no legal force, creates no offences, and confers no powers, and any account that describes its provisions as Indian law commits the precise error this series exists to correct.

What the draft would change, if enacted in the form circulated, can be stated shortly because the operative differences are few and structural. Devices would stand in the statute on their own footing, with their own advisory board, rather than inside the definition of a drug. Clinical trials and their compensation code would move from rules into the Act. The central regulator would acquire the statutory personality the CDSCO lacks. Licensing of manufacture would be substantially centralised, adopting the verdict side of the federal debate that the present article has resisted in part, which is one reason state governments and small manufacturers have been among the draft’s most attentive critics. And online sale of medicines, conducted for years in the interstices of a statute written before the internet, would be brought within an express licensing chapter. The device industry’s objection, voiced by the Association of Indian Medical Device Industry in its letter seeking the draft’s recall, adds a further fork: not consolidation at all, but separation, a devices law of their own, for which a recommendation of the Parliamentary Standing Committee on Health is cited in support. Between a consolidated bill nobody has introduced and a separate bill nobody has drafted, the sector continues to be governed by the notification of February 2020.

Nor is the device industry the only constituency the draft unsettled, and the pattern of the objections explains the bill’s paralysis better than any drafting defect. The proposal to centralise manufacturing licensing alarmed the state governments, whose drug control organisations administer the present system and whose assent a centralising statute does not legally need but practically courts; it alarmed small and medium manufacturers, who feared a distant single licensing window, compliance costs pitched at the scale of the largest firms, and the loss of the state-level relationships through which their licences have always moved; and it left the retail trade, whose lakhs of licensed chemists the draft’s online pharmacy chapter would place beside a new class of licensed electronic competitors, divided between established chains and neighbourhood shops. A replacement statute in this field must carry the Centre, the states, the largest export industry in Indian manufacturing and the smallest shopkeeper in the same majority, and the draft’s decade in limbo is the record of that coalition failing to assemble. The 1940 Act survives its own obituary because every interest that denounces it can imagine a successor it likes less.

The series thesis can therefore be closed on its strongest instance. The Drugs and Cosmetics Act 1940 survives into its ninth decade not because its text answers the questions put to it, but because its schedules, rules and notifications answer them in its name, and because the institution that would replace it has been designed twice and enacted never. For the manufacturer, the operative law is Schedule M and the conditions on a state licence; for the sponsor, the New Drugs and Clinical Trials Rules; for the chemist, Schedules H, H1 and X; for the prosecutor, Sections 17A, 17B and 27 and the discipline of Section 25; for the patient, the categories that decide what a failed medicine legally is. The citizen’s part in the machine is not negligible either: a suspect medicine can be reported to the state drug controller, an informant’s information about the movement of spurious drugs is sought under the Centre’s whistle-blower scheme with rewards for information that leads to seizures, and anyone assembling the purchase bills, photographs and correspondence such a report needs has reason to keep that dossier ordered and private, as the account of how journalists use VaultBook to protect sources shows in describing the encrypted single-file workspace as a tool for holding sensitive material together. The Act that governs every medicine in India is, in the end, best understood not as the short text of 1940 but as the whole apparatus the text set in motion: approved once at the Centre, licensed state by state, tested in laboratories the Act created, argued about in committees the Act requires to be consulted, and awaiting a successor that exists, complete and unenacted, in a draft.

How to read this statute

A reader who has reached this point owns the machine whole, and the article closes with the reading method its parts imply, because the Act punishes the reader who consults only the parent text. Three habits serve. The first is to ask, of any concrete question, which instrument answers it: the Act for the offence and the power, the Rules for the procedure and the condition, the schedules for the standard and the category, and the gazette notification for the perimeter, since what counts as a drug, a device or a banned combination on a given date is a fact of notification. The second is to ask which half of the machine is speaking: a CDSCO circular and a state licensing order are not interchangeable documents, and the citizen’s remedy, the trader’s compliance duty and the prosecutor’s evidence all differ with the speaker. The third is to ask which of the statute’s strata is in view, the 1940 import code, the 1982 penal reconstruction, the 2008 process reforms, or the decriminalising trims of the Jan Vishwas exercise, because the text on the page is all four at once and no single Parliament ever approved the composite. The replacement bill, if a future Parliament takes it up, will be judged against exactly the machine described here: one text, two administrations, a pharmacopoeia for a standard, an inspectorate for a spine, and a market that tests every joint of it daily.

For the three readers this series serves, the method cashes out differently. The patient or the patient’s family meets the Act at its hardest edge, and the route runs from the preserved strip and invoice, through the state drug control organisation’s complaint channel and the adverse-event reporting of the pharmacovigilance programme, to the person-aggrieved complaint that Section 32 permits and the consumer remedies that run in parallel; the article’s counter section and the traceable chain are the provisions that will do the work. The trader and the manufacturer meet the Act as a file: the licence, its conditions, the schedules governing their category, and the retention and inspection cycle, and their discipline is documentary, because every defence the Act gives them, warranty, due diligence, controversion of the analyst’s report, is a paper defence with a deadline. The student and the civil service aspirant meet the Act as a map of Indian federalism in working order, and should hold the cluster together: this statute for quality, the medical councils’ successor for the profession, the essential commodities code for price, and the patent statute for property, four instruments that public argument constantly fuses and that the examination, like the court, insists on keeping apart. That discipline of separation, applied to one statute read whole, is the habit this article has tried to model: text, rules, schedules and notifications in their places, each administration in its own, and every number attributed to the source that counted it.

Frequently Asked Questions

Q: What does the Drugs and Cosmetics Act regulate?

The Act regulates the import, manufacture, distribution and sale of drugs and cosmetics in India. Its definition of a drug is deliberately wide, covering medicines for internal or external use in human beings and animals, substances used in diagnosis, treatment, mitigation or prevention of disease, substances intended for use as components of a drug including empty gelatin capsules, and medical devices notified by the Central Government. Through the Drugs and Cosmetics Rules 1945 it controls who may make or sell these products, on what premises, under whose supervision, to what standard, and with what records. It also governs clinical trials and new drug approvals through the New Drugs and Clinical Trials Rules 2019.

Q: What is a spurious drug and what is the punishment?

A drug is spurious under Section 17B if it is manufactured under a name belonging to another drug, imitates or substitutes another drug in a manner likely to deceive, bears the name of a fictitious or non-existent manufacturer, has been substituted wholly or partly by another substance, or claims to be the product of a manufacturer who did not make it. The category is about false identity, not weakness. Punishment under Section 27 depends on harm: a spurious drug likely to cause death or grievous hurt draws imprisonment of not less than ten years extending to life, with heavy fines directed to the victim as compensation; otherwise, not less than seven years extending to life.

Q: Who approves a new drug in India?

A new drug is approved at the Centre by the Central Licensing Authority, which in practice is the Drugs Controller General of India, heading the Central Drugs Standard Control Organisation. The approval covers permission to import the drug or to manufacture it for sale, and is granted under the New Drugs and Clinical Trials Rules 2019 on the strength of safety and efficacy data, including clinical trial evidence where required. No state authority can approve a new drug, and no state manufacturing licence can lawfully be used to market a product that lacks central approval as a new drug. Approval is a judgment about safety and efficacy; it confers no patent or exclusivity rights.

Q: Who licenses a pharmacy or a manufacturing unit?

The licensing authority of the state or union territory in which the premises are located. Manufacturing licences and sale licences, wholesale and retail, are all granted by state licensing authorities under the Drugs and Cosmetics Rules, which also set the premises, staffing and record conditions attached to them. A pharmacy trading in three states holds licences from three authorities. One qualification exists at the top of the risk range: for categories such as blood products, vaccines, sera, large-volume parenterals and recombinant DNA products, the state licence follows approval by the Central Licence Approving Authority at the Centre. Central approval of a new drug and state licensing of the premises that make and sell it are separate acts by separate governments.

Q: Are medical devices regulated as drugs?

Yes, by the statute’s own device. Section 3(b)(iv) of the Act treats as drugs such devices as the Central Government notifies after consulting the Drugs Technical Advisory Board. For decades only a small set of categories was notified; a notification of 11 February 2020, S.O. 648(E), then specified all medical devices, from syringes and implants to software and accessories, as drugs with effect from 1 April 2020. The operative regime is the Medical Devices Rules 2017, which classify devices by risk into Classes A to D and license them proportionately. The arrangement is widely regarded as a borrowed fit, and a separate devices statute has been recommended, but the drugs statute is where device regulation legally lives.

Q: What are the schedules under the drug rules?

The schedules to the Drugs and Cosmetics Rules 1945 carry most of the operative law. Schedule M prescribes good manufacturing practices for factories. Schedules H, H1 and X list prescription medicines of rising strictness: H for general prescription sale, H1 for named antibiotics, anti-tuberculosis and habit-forming drugs sold against prescription with a three-year register, and X for dependence-producing psychotropics requiring duplicate prescriptions and special licences. Other schedules fix standards and shelf life (F and P), biological products (C and C1), the caution-labelled medicines (G), diseases no drug may claim to cure (J), and the records licensees must keep (U). Schedule Y, which governed clinical trials, was displaced by the New Drugs and Clinical Trials Rules 2019.

Q: How are clinical trials regulated in India?

Clinical trials are governed by the New Drugs and Clinical Trials Rules 2019, which replaced the older Schedule Y regime. A trial requires permission from the Central Licensing Authority and approval by a registered ethics committee at the trial site, together with the informed consent of each participant. The Rules commit the Authority to decide a trial application within ninety working days, or thirty working days for a drug discovered or proposed to be manufactured in India, after which silence counts as permission. They prescribe free medical management for injured participants and regulator-fixed compensation for trial-related injury or death, and they allow waiver of a local trial for certain drugs approved in specified jurisdictions.

Q: Is the Drugs and Cosmetics Act being replaced?

A replacement exists in draft but has not been enacted. The Ministry of Health and Family Welfare published the draft Drugs, Medical Devices and Cosmetics Bill in July 2022 for public comment, proposing a consolidated statute that would cover devices and clinical trials expressly, give the central regulator a statutory foundation, centralise much manufacturing licensing and license online pharmacies. A version was listed for introduction in Parliament during the monsoon session of 2023 and was not introduced. An earlier amendment bill of 2013 was withdrawn in 2016. Until Parliament enacts a successor, the 1940 Act, its 1945 Rules and the notifications under them remain the whole of the law in this field.

Q: What is the difference between a spurious drug and a drug that is not of standard quality?

The difference is between lying about identity and failing a standard. A drug is not of standard quality when it fails the pharmacopoeial standard that Section 16 and the Second Schedule prescribe, because its content assays low, or it fails a dissolution test, or it has degraded; the conclusion says nothing about who made it or what it pretends to be. A drug is spurious under Section 17B when its identity is false: it imitates another product, bears a fictitious manufacturer’s name, or claims a maker who did not make it, and it is spurious even if its contents are of full strength. The legal consequences differ sharply, since spurious offences carry the ladder’s highest punishments, cognizability and restricted bail, while quality failures are often met with licence action.

Q: What compensation is payable when a clinical trial participant is injured or dies?

The New Drugs and Clinical Trials Rules 2019 make the sponsor pay on a no-fault basis. An injured participant must first receive free medical management for as long as the injury requires, or until the injury is shown to be unrelated to the trial. Where injury is trial-related, financial compensation is fixed by the Central Licensing Authority on the recommendation of an expert committee, computed under the formula the Rules prescribe, which takes account of the participant’s age and the seriousness of the injury or disability. Where the participant dies, compensation computed under the Rules is payable to the nominee or legal heir. The amount is set by the regulator, not negotiated with the sponsor, and proof of negligence is not a condition of payment.

Q: What is Schedule H1 and why must a chemist keep a separate register for these drugs?

Schedule H1 is a list within the Drugs and Cosmetics Rules, inserted in 2013 and in force from 1 March 2014, covering named higher-generation antibiotics, anti-tuberculosis medicines and habit-forming drugs, forty-six substances at its creation. Its purpose is to fight antimicrobial resistance and casual self-medication by attaching an audit trail to medicines that the ordinary prescription rule had failed to control. A chemist supplying a Schedule H1 drug must record the sale in a separate register showing the prescriber’s name and address, the patient’s name, the drug and the quantity, keep the register for three years open to inspection, and sell only against a registered medical practitioner’s prescription. The pack bears a red-bordered warning box and a red Rx symbol.

Q: Can medicines legally be sold online in India?

Only through the existing licensing structure, because the Act contains no chapter on electronic commerce. A seller of medicines must hold the sale licences the Rules require for the premises from which the medicines are stocked and dispatched, and prescription medicines may be supplied only against a valid prescription, whether the order arrives across a counter or through a website. Draft rules to regulate online pharmacies expressly were published for comment in 2018 and were never finalised, and the draft replacement bill of 2022 proposed an express licensing regime for online sale, which remains unenacted. The result is that online sale is lawful where the seller is licensed and the prescription rules are obeyed, and unlawful where either condition fails, but the dedicated regulatory code for the sector does not yet exist in force.

Q: Who can start a prosecution for an offence under the Act?

Section 32 restricts the gate. A prosecution under Chapter IV may be instituted only by a Drugs Inspector, by a gazetted officer of the Central Government or a State Government authorised in writing for the purpose, by the person aggrieved, or by a recognised consumer association, whether or not the person aggrieved is a member of it. The police have no general power to launch a prosecution under the Act by themselves, although offences relating to adulterated and spurious drugs are cognizable under Section 36AC, so police may investigate them. In practice almost all prosecutions are inspector-led, founded on a drawn sample and a Government Analyst’s report, and since the amendment of 2008 the offences are tried by a Court of Session, with designated Special Courts for the adulterated and spurious categories.

Q: What is the Indian Pharmacopoeia and how does it fix the standard a drug must meet?

The Indian Pharmacopoeia is the official book of drug standards published by the Indian Pharmacopoeia Commission, an autonomous body of the Ministry of Health and Family Welfare. Section 16 of the Act defines standard quality for a drug by reference to the Second Schedule, and the Second Schedule points to the Pharmacopoeia as the governing standard for the products it covers, with other named pharmacopoeias filling the gaps for products it does not. Each monograph specifies the identity, purity, strength and test methods for a medicine, so a laboratory asked whether a sample is of standard quality is really asked whether the sample passes the Pharmacopoeia’s tests. A failure is the legal basis of the not of standard quality finding on which most enforcement rests.

Q: How are Ayurvedic, Siddha and Unani medicines regulated under the Act?

They are regulated inside the Act but in a chapter of their own. Chapter IVA, inserted in 1964, governs Ayurvedic, Siddha and Unani drugs, with its own advisory board, its own definitions of misbranding and adulteration adapted to traditional formulations, and its own penalty provisions. Manufacture requires a licence from the state licensing authority, and the good manufacturing practices for these products are set out in Schedule T to the Rules, which prescribes premises, hygiene and record requirements fitted to traditional manufacture. The standards reference point is the authoritative books of the systems listed in the First Schedule to the Act and the Ayurvedic Pharmacopoeia. These medicines are not exempt from the Act; they are governed by a parallel code within it.

Q: What is a fixed-dose combination and why have so many been banned in India?

A fixed-dose combination puts two or more active ingredients in a single tablet, capsule or injection. India accumulated thousands of such combinations, many licensed by individual states without the central approval that a new drug requires, and many combining ingredients with no therapeutic justification for the pairing. Section 26A lets the Centre prohibit a drug that involves risk or lacks the therapeutic value claimed for it. Acting on expert committee advice, the Centre prohibited 344 combinations in March 2016; the Delhi High Court struck the bans down that December for bypassing the Drugs Technical Advisory Board; the Supreme Court referred the combinations to the Board in December 2017; and on its report the Centre prohibited 328 combinations in September 2018.

Q: What should a patient do if a medicine they have bought looks fake or makes them ill?

The complaint goes to the enforcement tier of the patient’s own state: the state drugs control organisation, whose inspectors can draw samples of the suspect product, send them for analysis and act against the seller’s licence. Keeping the strip, the packet and the purchase bill matters, because a sample is worth little without its source. A purchaser also has a statutory route of their own under Section 26, which entitles any person to have a drug analysed by a Government Analyst on payment of the prescribed fee and to receive the report. Serious reactions to a genuine medicine are a different problem and can be reported through the Pharmacovigilance Programme of India, which collects adverse drug reaction reports from patients and hospitals. Detail here is general information, not advice for any individual case.

Q: For how long does an approved drug remain a new drug?

The answer changed with the 2019 Rules. Under the earlier Rule 122E of the Drugs and Cosmetics Rules, an approved drug continued to be treated as a new drug for four years from its approval, and during that period manufacture and marketing remained tied to the central approval. The New Drugs and Clinical Trials Rules 2019 abandoned the fixed period and define a new drug functionally: broadly, a drug that has not been used in the country to any significant extent, is not already approved and marketed for the claims proposed, or is proposed in a new combination, indication, dosage form or route of administration. On the functional test, a product’s newness is a question about its history in the Indian market rather than about the calendar.

Q: What is the Drugs Technical Advisory Board and why must it be consulted before the rules change?

The Board is the expert body the Act itself constitutes in Section 5, chaired by the Director General of Health Services and composed of officials, pharmacologists, clinicians, pharmaceutical scientists and representatives of the industry and the profession. Its function is to advise the Central and state governments on technical matters arising from the administration of the Act. The consultation matters legally because Sections 12 and 33 make it a condition of the rule-making power: before the Central Government makes rules under the import or manufacture chapters, it must consult the Board and publish the draft for objections. Rules made without that consultation are open to challenge on that ground alone, and the litigation over the 2016 fixed-dose combination bans, in which the Delhi High Court fastened on the absence of expert consultation before the Supreme Court routed the combinations through the Board for examination, shows how heavily the Act’s legitimacy machinery depends on the Board being used rather than bypassed.

Q: What happens when a manufacturer’s licence is suspended or cancelled, and can the order be appealed?

Suspension and cancellation are the licensing authority’s own sanctions, imposed under the conditions of the licence and the Drugs and Cosmetics Rules rather than by a criminal court. Where an inspection or a failed sample shows a contravention, the authority issues a show cause notice to the licensee, considers the reply, and may suspend the licence for a stated period or cancel it, with a stop-production direction where manufacture must halt at once. The sanction is administrative, so it operates without waiting for a prosecution to conclude, which is why it is the consequence manufacturers encounter most often. The Rules give the aggrieved licensee an appeal to the state government against the licensing authority’s order, and constitutional remedies in the High Court remain available beyond the statutory appeal.