Criminal law reform is the subject on which well-prepared candidates lose marks they did not need to lose. The candidate who understood murder, arrest and the burden of proof perfectly well under the old codes opens a paper that cites the new ones, reaches for a section number that has moved, and writes a confident answer about the wrong provision. The knowledge was there. The retrieval failed. That distinction, between not knowing the law and not being able to find it under examination pressure, is the whole subject of this guide, because the replacement of the Indian Penal Code, the Code of Criminal Procedure and the Indian Evidence Act by the Bharatiya Nyaya Sanhita, the Bharatiya Nagarik Suraksha Sanhita and the Bharatiya Sakshya Adhiniyam changed the map of Indian criminal law far more than it changed the territory.
The method defended here can be stated in one sentence: hold every offence and doctrine by its elements and its chapter position, and treat both numbering systems as lookups against that structure. A candidate who knows that murder is the aggravated form of culpable homicide, sitting in the part of the code that deals with offences against the human body, next to attempt and abetment of suicide and causing death by negligence, can find the provision in either code in seconds. A candidate who memorised a bare list of numbers owns a list that the renumbering has just halved in value. Everything that follows is an unpacking of that sentence: the three-code spine, the four moments that organise the field, the lattice that joins the two systems row by row, the small set of numbers that function as names and must be carried in the head, the judgments whose propositions carry the marks, and the transitional rule that examiners can test from either direction.
Three codes, three questions: knowing which code a question belongs to
Indian criminal law is not one subject but three, and the first discipline is knowing which of the three a question is really about. The substantive code defines offences and punishments: it answers the question what has been done and what the state may do about it. The procedure code governs everything that happens between the first information and the final appeal: it answers the question how the machinery moves, who may set it in motion, and what restraints bind it at each stage. The evidence code governs proof: it answers the question what a court may look at, who must establish it, and to what standard. A remarkable share of lost marks in criminal law papers comes from answering a question drawn from one code with material drawn from another, writing about the definition of an offence when the question asked about the stage at which it must be proved, or reciting the punishment when the question asked about the forum.
The three codes were replaced as a set, and the replacements preserve the division exactly. The Bharatiya Nyaya Sanhita took over the substantive territory of the Indian Penal Code. The Bharatiya Nagarik Suraksha Sanhita took over the procedural territory of the Code of Criminal Procedure. The Bharatiya Sakshya Adhiniyam took over the evidentiary territory of the Indian Evidence Act. A candidate who can sort any question into substance, procedure or proof within seconds has already avoided the commonest structural error, and the sorting survives renumbering because it depends on the function of the provision rather than its address. When a question asks whether a statement can be proved, the answer lives in the evidence code whatever number the provision wears. When it asks who may order an investigation, the answer lives in the procedure code. When it asks what must be intended or known before liability attaches, the answer lives in the substantive code.
Which of the three codes decides whether a confession to a police officer can be used?
The evidence code decides it, in both systems. The bar is a rule about what a court may look at, not about how the police investigate, and it sits in the evidence statute: Section 25 of the Indian Evidence Act, carried into Section 23 of the Bharatiya Sakshya Adhiniyam, which gathers the old Sections 25 to 27 into one provision.
That single example teaches the general habit. Confessions feel like investigation, and investigation feels like procedure, so candidates reach for the procedure code and find nothing, because the rule was never there. The investigation provisions tell the police what they may record; the evidence provisions tell the court what it may hear. Sorting by function rather than by atmosphere is the skill, and it is worth practising deliberately: take any provision from either set of codes and state, in one clause, which of the three questions it answers. A candidate who can do that for the fifty provisions that dominate the syllabus has built the shelf on which everything else in this guide is stored. The shelf is also the reason the transition is less frightening than it looks, because the three-question structure of 1860, 1872 and 1973 is the three-question structure of the successor codes, and a candidate who owns the structure owns the frame into which every renumbered provision can be placed on first sight.
The four moments that organise criminal law reform
A candidate does not need a chronology of Indian criminal law. A candidate needs four moments, held firmly, because almost every provision and judgment in the syllabus can be placed in relation to one of them, and a fact that can be placed can be retrieved. The four moments are the codification project of the nineteenth century, the procedural reconstruction of 1973, the sexual offences reform of 2013, and the replacement of the codes in 2023 and 2024. Everything the examiner can ask hangs from one of those four pegs.
The first moment runs from 1834, when the First Law Commission under Thomas Babington Macaulay began work, through the draft Penal Code of 1837, to the enactment of the Indian Penal Code on 6 October 1860 and its commencement on 1 January 1862. The Indian Evidence Act followed in 1872, drafted principally by James Fitzjames Stephen and in force from 1 September of that year. What a candidate should carry from this moment is not the dates for their own sake but the design philosophy: codification was a deliberate break with a patchwork of regulations and personal laws, an attempt to state the criminal law of a vast territory in a single systematic text arranged by the nature of the offence. That is why the Penal Code is organised by chapters of offences rather than by the order in which a case unfolds, and why its chapter architecture, general explanations and general exceptions first, then offences grouped by what they attack, became the template the successor code largely kept. The colonial origin of the codes is a fact about their birth, not an argument about their content, and answers that treat it as an argument score poorly against answers that can say what the codifiers were trying to build.
The second moment is the Code of Criminal Procedure of 1973, enacted as Act 2 of 1974 and in force from 1 April 1974. Its organising act was the separation of the judiciary from the executive at the district level: judicial magistrates, who try cases, were placed under the High Court, and executive magistrates, who keep the peace and handle preventive functions, under the state government. That separation, implementing the constitutional directive in Article 50, is the reason the procedure code distinguishes two species of magistrate, and a candidate who understands the reason can reconstruct which magistrate does what without memorising the allocation. The 1973 code also carried the arrest, bail and remand architecture that the courts spent the next five decades disciplining, which is why so many of the syllabus judgments are arguments about a code that gave the police wide powers and left the courts to build the restraints.
The third moment is the Criminal Law (Amendment) Act 2013, Act 13 of 2013, the legislative answer to the Delhi gang rape of December 2012 and to the report of the Justice Verma Committee, which was constituted on 23 December 2012 and reported within twenty-nine days. An ordinance came first, on 3 February 2013, and the Act was deemed to have come into force from that date. The amendment widened the definition of rape, created the graded offences of sexual harassment, disrobing, voyeurism and stalking in Sections 354A to 354D of the Penal Code, raised punishments, and changed the evidence rules for sexual offence trials. For examination purposes the moment matters twice over: it is the substantive reform most likely to be asked about in its own right, and it is the clearest example in the whole field of how reform actually happens, by public shock, committee report, ordinance and Act in that order, with some committee recommendations accepted and others, most famously the criminalisation of marital rape and the rejection of the death penalty for rape, refused.
The fourth moment is the replacement of 2023 and 2024. The three new codes were passed by Parliament in December 2023, received the President’s assent on 25 December 2023, and came into force on 1 July 2024: the Bharatiya Nyaya Sanhita as Act 45 of 2023, the Bharatiya Nagarik Suraksha Sanhita as Act 46 of 2023, and the Bharatiya Sakshya Adhiniyam as Act 47 of 2023. The Penal Code’s 511 sections were reorganised into 358; the procedure code’s 484 sections into 531; the Evidence Act’s 167 sections into 170. Most of the substantive law carried over. What changed was the numbering, a defined set of substantive choices, and the names. A candidate holding the four moments can place any provision: a rule about the separation of magistrates belongs to 1973, a graded sexual offence to 2013, a renumbered offence with a new name to 2024, and the chapter logic beneath all of it to 1860 and 1872.
Elements before numbers: the organising method
The method itself deserves a section, because it is the claim on which this guide stands or falls. An offence is not a number. An offence is a set of elements: the conduct required, the circumstance in which it must occur, the result it must cause where a result is required, and the state of mind, intention, knowledge, dishonesty, rashness or negligence, with which it must be done. Murder is not Section 302 of the old code or Section 103 of the new one. Murder is culpable homicide committed with one of the specified aggravated mental states, and the definition, the exceptions that pull a killing back down to culpable homicide not amounting to murder, and the punishment live in that structure. A candidate who holds the structure can be shown either number and know what is being asked. A candidate who holds only the number can be shown the other code’s number and know nothing.
Chapter position is the second half of the method, and it works because the successor code kept the old one’s architecture while improving its order. In both substantive codes the general provisions come first: definitions, the rules on common intention and abetment, the general exceptions that excuse or justify. The offences then follow grouped by their object. The Bharatiya Nyaya Sanhita improved one piece of the ordering that examiners have noticed: the offences against women and children, which the Penal Code scattered between the chapter on the human body and the chapter on marriage, are gathered into a single chapter placed near the front of the code, sections 63 to 99, so that rape, the offences built around it, cruelty by a husband or his relatives at Section 85 and dowry death at Section 80 sit within one stretch of text. The candidate who knows that the women’s offences live together near the front, that the human body offences follow, and that property offences come later in the code, can navigate an unfamiliar number by asking where in the building it must be.
Why does chapter position survive renumbering when a section number does not?
Because renumbering changed the addresses and kept the building. The successor code preserved the habit of grouping offences by what they attack, principles first and specific offences after, so a candidate who knows which room a doctrine lives in can find its number from the neighbourhood, while a memorised number says nothing about its neighbours.
The practical consequence is a change in what revision time buys. An hour spent memorising fifty numbers buys fifty fragile facts, each of which the other code falsifies. An hour spent on the elements of ten offences and their chapter positions buys a structure that generates the answers to questions the candidate has never seen, including the deliberately awkward questions, of which examiners are fond, that cite a provision by description rather than by number. This is also why the method scales across all three codes. The procedure code is a sequence, investigation, arrest, remand, charge, trial, judgment, appeal, and a candidate who holds the sequence can place a procedural provision by the stage it governs. The evidence code is a logic, relevancy first, then proof, then burden and presumptions, and a candidate who holds the logic can place an evidence provision by the question it answers. Elements and position, sequence and logic: the three codes have three shapes, and the shapes did not change in 2024.
The architecture of the substantive code as a finding system
The elements method works because the substantive code is built as a finding system, and a candidate who internalises its floor plan can locate a provision that has never been memorised. The Sanhita opens with its general provisions, and the front of the code is where the machinery of all liability lives: the definitions, the rules on territorial reach, the general explanations that apply across every chapter, and the provisions on joint liability, including the common intention rule at Section 3(5), which the Penal Code kept at Section 34. Behind the machinery come the punishments, and then the general exceptions, the provisions that excuse or justify what would otherwise be offences: acts done by a person bound or justified by law, acts done in good faith, acts done under compulsion, and private defence, which the code treats as a structured doctrine with its own commencement and continuance rules rather than as a slogan. A candidate should learn the front of the code as a unit, because examiners draw on it constantly and because its logic, liability first requires a prohibited act done with a culpable mind unless an exception or a defence intervenes, is the grammar of every substantive question.
Then come the offences, grouped by what they attack, and the grouping is the finding system. Offences against women and children come first in the Sanhita among the offence chapters, gathered in a single chapter running from Section 63 to Section 99, a deliberate act of arrangement that a candidate should be able to describe and evaluate, because the Penal Code had scattered those offences across its length and the consolidation is itself an examinable design decision. Offences affecting the human body follow, and within that neighbourhood the candidate should be able to walk from homicide to hurt to wrongful restraint and confinement as a connected street rather than as separate addresses. Offences against property form their own neighbourhood, where theft, extortion, robbery, cheating and criminal breach of trust sit in a graded family, and offences against the state, public tranquillity, public health and public justice occupy theirs. The walk matters more than the addresses. A question that describes conduct without naming it is answered by identifying the victim and the interest attacked, walking to the neighbourhood, and then distinguishing among the graded offences inside it by their elements, which is precisely the skill the elements column of the lattice trains.
The general exceptions deserve a paragraph of their own, because they are the most under-revised high-yield material in the substantive code. Private defence is tested through its limits: the right extends to causing death only in the situations the code enumerates, it commences when reasonable apprehension commences and continues while the apprehension continues, and it is unavailable where there is time to have recourse to public authorities. Mistake of fact, accident, infancy, insanity and intoxication each turn on a definitional hinge that a problem question will press. The candidate who learns the exceptions as a set of hinges, each with the precise word or phrase on which liability turns, can answer problem questions across the whole offence range, because examiners build problems by taking a standard offence and inserting an exception-shaped fact. That is also why the exceptions carry over through the transition almost untouched: they state the conditions of responsibility itself, and no renumbering reaches them.
The procedural sequence as a finding system
Procedure is learned badly when it is learned as a list of powers and well when it is learned as a journey with stages, because the code itself is arranged as a journey. The sequence begins with information: the first information report at Section 173 of the procedure Sanhita, the provision that converts a complaint into a state process, with its rules on recording, on reading the information over to the informant, on the informant’s copy, and, in the new code, on information given electronically, which must be taken on record and signed by the informant within three days. Investigation follows, with the police powers to record statements, to search, to seize, and to arrest under Section 35, where the necessity test and the notice of appearance govern. The remand provisions at Section 187 organise the investigative clock, the first production before a magistrate, the fifteen days of police custody the code makes available across the opening stretch of the investigation, and the sixty and ninety day ceilings at whose expiry bail becomes a right. Cognizance, complaint and committal form the bridge into court; charge, trial in its several forms, evidence, argument and judgment form the trial proper; and appeal, revision and the inherent powers close the system.
Each stage carries its own examinable doctrine, and the doctrines are attached to stages rather than to numbers. At the information stage the doctrine is Lalita Kumari, the mandatory registration rule with its narrow preliminary inquiry window, to which the new code adds the possibility of a preliminary enquiry, before registration, for offences punishable with imprisonment between three and seven years, itself a provision that invites comparison with the judgment. At the arrest stage the doctrines are necessity and notice, Arnesh Kumar and the codified checklist. At the remand stage the doctrine is the entitlement to bail on expiry of the investigation period, a right the courts describe as indefeasible. At the trial stage the doctrines are the charge, which must state the offence and give the accused notice of the matter charged, and the separation of functions between prosecution and court. A candidate who binds each doctrine to its stage can reconstruct an answer to a staged problem question in order, stage by stage, even where the precise provision number has to be supplied from the lattice rather than from memory.
The bail cluster rewards separate treatment, because bail questions are a standing genre and the transition rearranged the cluster’s addresses. Bail in bailable offences sits at Section 478 of the new code, bail in non-bailable offences at Section 480, special powers of the High Court and the Sessions Court at Section 483, anticipatory bail at Section 482, and the undertrial detention ceiling, the successor of the old Section 436A, at Section 479, which in the new code extends a measure of its relief to first-time offenders earlier in their detention. The old code’s arrangement, Sections 436 to 439 with 436A and 438, was learned by generations as a single neighbourhood, and the candidate’s task is not to unlearn the neighbourhood but to re-map its internal addresses while keeping its logic: the distinction between bail available by right and bail granted in the court’s discretion, the conditions a court may impose, and cancellation, are stable doctrines wearing new numbers.
The logic of the evidence code as a finding system
The evidence code is the shortest of the three and the most purely logical, and it repays being learned as a single argument. The argument runs in three movements. The first movement asks what is relevant: facts in issue and the facts the code declares relevant to them, res gestae at the front of the scheme in the old Section 6 and the new Section 4, motive, preparation and conduct, admissions and confessions in their carefully separated compartments, statements by persons who cannot be called, including dying declarations, and the opinions of experts, which the code treats as relevant but weak material that assists the court rather than binds it. The second movement asks what must be proved and by whom: the burden provisions that open with the general rule at the old Section 101 and the new Section 104, and the presumptions, where the code tells the court it may presume, shall presume, or must treat a matter as conclusive proof, three expressions whose difference is a standing examination point because each carries a different degree of compulsion. The third movement asks what may be looked at and in what form: documentary evidence, the primary and secondary distinction, the special regime for electronic records, and the exclusionary rules, of which the confession provisions are the most heavily tested.
Two distinctions inside that argument carry a disproportionate share of the marks. The first is between an admission and a confession: every confession is an admission, but an admission is the wider genus, made in civil and criminal contexts alike, while a confession is a statement by an accused admitting the offence or substantially all the facts constituting it, and the code’s exclusions, the bar on confessions to a police officer and the bar on confessions made in police custody outside a magistrate’s immediate presence, attach to confessions specifically. The second is the architecture of the custody confession rule itself: a confession made in custody is unusable, except that so much of the information given by an accused in custody as relates distinctly to a fact discovered in consequence of it may be proved, the discovery doctrine that generations of students have learned through its leading illustration of the weapon recovered at the accused’s instance. Both distinctions survive the transition entire, restated in the consolidated Section 23 of the Adhiniyam, and both are best learned as pieces of reasoning rather than as text: the reason for excluding custodial confessions is the unreliability produced by the setting, and the reason for the discovery exception is that a fact independently verified by discovery carries its own guarantee of truth.
Electronic evidence deserves its own place in the evidence revision plan, because it is the one area where the new code visibly adds machinery. The old law located the certificate requirement in Section 65B of the Evidence Act, inserted in 2000, and the courts built the doctrine in Anvar and Arjun Panditrao. The Adhiniyam restates the scheme at Section 63 and prescribes the certificate in its Schedule, in two parts, one for the person producing the record and one including an expert’s contribution, with the hash value of the record among the particulars the form calls for. The examinable material is therefore three-layered: the statutory scheme, the case law that governs its operation, and the question of how the new formality interacts with the case law, a question on which the honest study answer is that the certificate jurisprudence continues to govern where the provision is substantially the same, and the prescribed form fixes the certificate’s content in a way the old provision left to practice.
The judgments method: propositions, courts and confidence
Because marks follow propositions, candidates need a method for propositions as much as a list of them. The method has three rules. The first rule is to state what the court decided, in the indicative mood, without the facts unless the facts are the point: Maneka Gandhi held that a procedure under Article 21 must be just, fair and reasonable; the passport facts are retrieval colour, not the proposition. The second rule is to know the weight of the decision: a Constitution Bench decision such as Lalita Kumari binds differently from a two-judge decision, and where two decisions interact, the candidate should know which governs, as Arjun Panditrao governs the timing of the certificate because it reconsidered and restated the position after Anvar and after the brief relaxation in Shafhi Mohammad. Examiners probe exactly this ordering when they ask a candidate to reconcile apparently conflicting authorities, and the answer that states the hierarchy and the chronology, and then applies the governing proposition, is the answer that scores. The third rule is candour about boundaries: a proposition has an edge, and the strong answer states where the holding stops, as with the preliminary inquiry in Lalita Kumari, which the judgment permits only in limited categories and only to test whether a cognizable offence is disclosed, not to test the truth of the information.
The constitutional frame deserves the same proposition discipline. Gopalan’s compartment theory and Maneka’s interrelation holding are propositions, and so is the proposition that Article 20(1) bars ex post facto criminal legislation while Article 20(2) bars a second prosecution and punishment for the same offence and Article 20(3) protects an accused against compelled self-incrimination; candidates confuse the three clauses of Article 20 with one another under pressure, and the confusion is expensive because each clause answers a different question. Article 21 supplies the fairness frame, Article 22 the safeguards on arrest and the whole preventive detention architecture, and a candidate who can assign any liberty problem to its clause in one sentence, and cite the governing proposition in the next, has the constitutional shelf built to the same standard as the three code shelves. The detention cluster inside that shelf, with its own statutes and its own examination tradition, is treated in the civil liberties examination guide named later in this article, and the two guides are designed to interlock rather than to overlap.
There is also a discipline for citing judgments whose details a candidate is unsure of, and it should be stated plainly because it protects marks. Where the proposition is certain and the citation is not, state the proposition and anchor it to the court and the subject matter in general terms rather than inventing a citation, a year or a bench strength. Where the proposition itself is uncertain, do not state it as a holding at all. Invented citations are the one unforgivable currency in this field: examiners forgive a missing reference, because the proposition carries the reasoning, and they do not forgive a confident falsehood, because it advertises that the candidate’s other propositions may also be fabricated. The same rule protects the candidate in the viva and in the interview, where a follow-up question will find a fabricated citation within two exchanges.
Reading old judgments in a new-code world
A transition creates a reading problem that candidates rarely prepare for: the entire library of authority was written under the old codes, and papers, courts and commentators will cite it in the old numbers for years. The candidate’s working rule is a principle of correspondence with a limit attached. Where a provision of a new code re-enacts a provision of an old code in substantially the same terms, decisions interpreting the old provision continue to govern the new one, because the legislature is taken to legislate against the settled interpretation of the words it chose to repeat. Where the new text changes the words in a way that matters, the authority weakens to the extent of the change, and the candidate must be able to say which case it is. Section 63 of the Adhiniyam is Section 65B re-enacted with a prescribed form added, so Anvar and Arjun Panditrao govern it, with the form question sitting on top of the case law rather than replacing it. The remand provision changed the structure of police custody, so custody decisions under the old Section 167 govern the new Section 187 only with that change marked. The sovereignty provision is not a re-enactment at all, so the sedition case law, Kedar Nath Singh’s reading of Section 124A, is background to the comparison rather than authority on the new offence. Three grades of inheritance, complete, qualified and none, and the candidate who grades the case law provision by provision can answer the sophisticated form of the carry-over question, which asks not whether the old law survives but where, and how far.
The reading skill that supports the grading is translation at sight. A judgment’s headnote will speak of Section 302 or Section 482 in the old senses; the candidate reads the number, attaches the doctrine from the elements and the context, and only then translates. Practised deliberately, with a page of any standard commentary open and each old citation converted aloud, the skill becomes fast enough for examination use within weeks, and it has a side benefit that candidates underrate: it reveals how thin the genuinely new ground is. Most pages of most commentaries translate without friction, sentence after sentence of doctrine surviving the crossing intact, because the doctrines were built on elements and reasons rather than on numbers. The pages that resist translation are the pages that matter, the custody restructuring, the electronic information rules, the new offences, the redrawn sovereignty provision, and translation practice is the most reliable instrument a candidate has for finding those pages in their own knowledge before an examiner finds them in the hall.
The interview and the viva: the transition spoken aloud
Judiciary and civil services candidates alike may meet the transition in spoken form, in a viva or an interview, and the spoken form punishes different weaknesses from the written one. A written answer can pause; a spoken answer must retrieve, and retrieval under observation exposes the candidate whose knowledge is a list, because a list can be recited but not searched. The interviewer who asks what changed in the law of remand is not asking for Section 187; the question is whether the candidate can describe a change structurally, the fifteen days of police custody made available in parts across the opening forty or sixty days instead of being confined to the first fifteen, and then say why it matters and what the courts will have to decide about it. Preparation for the spoken form is therefore production practice in the candidate’s own voice: stating propositions aloud, translating citations aloud, and giving the one-minute version of the transition, the codes, the date, the savings architecture, the finite list of real changes, until the account is a narrative the candidate owns rather than a paragraph the candidate has read. The candour discipline matters most here of all: an interviewer will pursue a confident error for the pleasure of the pursuit, while an honest boundary, stating that a point of detail would need the text, closes a line of questioning with credit intact.
The dual-code revision lattice
Everything argued so far becomes usable in one object: a lattice organised by offence and doctrine rather than by number, with the old section and the new section side by side, the elements to hold in the middle, and a warning wherever the two systems set a trap. The full translation exercise, provision by provision across the substantive codes, belongs to the IPC and BNS comparison, which owns that table for the series; the lattice here is a revision instrument, selective by design, covering the provisions that dominate examination papers across all three codes and built to be read in either direction, old to new when a judgment or a textbook cites the old code, new to old when a question paper cites the Sanhita and the candidate’s memory answers in the Penal Code.
| Offence or doctrine | Old section | New section | Elements to hold | Leading judgment | Confusion warning |
|---|---|---|---|---|---|
| Murder, punishment | IPC 302 | BNS 103 | Culpable homicide with an aggravated mental state from the definition section | None; hold the definition and its exceptions | The number 302 has moved; writing it for murder dates the answer instantly |
| Rape, definition and punishment | IPC 375 and 376 | BNS 63 and 64 | Absence of consent as the organising idea; the marital exception as read down | Independent Thought | The old text said fifteen; the courts read eighteen, and the new code carries eighteen |
| Cruelty by husband or relatives | IPC 498A | BNS 85, definition in 86 | Cruelty, including conduct linked to unlawful demands for property | None | The provision sits in the women and children chapter, not beside the body offences |
| Dowry death | IPC 304B | BNS 80 | Death within seven years of marriage, cruelty soon before death | None | The offence and its presumption live in different codes; see the presumption row |
| Causing death by negligence | IPC 304A | BNS 106 | A rash or negligent act not amounting to culpable homicide | None | Section 106(2) adds a separate, heavier hit-and-run limb |
| Attempt to murder | IPC 307 | BNS 109 | An act done with the intention or knowledge of murder, short of death | None | Hurt caused in the attempt raises the punishment within the same provision |
| Sedition and its successor | IPC 124A | BNS 152 | Exciting disaffection, against endangering sovereignty, unity and integrity | Kedar Nath Singh | Not a renumbering: the successor is a differently drawn offence, and sedition as a name is gone |
| Common intention | IPC 34 | BNS 3(5) | A criminal act done by several persons in furtherance of a shared intention | None | Moved into the general provisions at the front of the code |
| Criminal conspiracy | IPC 120B | BNS 61 | Agreement between two or more persons to do an illegal act or a legal act by illegal means | None | The definition sits at Section 61(1); the punishment follows it |
| Theft | IPC 378 and 379 | BNS 303 | Dishonest taking of movable property out of another’s possession without consent | None | Snatching is carved out as its own offence at Section 304 |
| Cheating | IPC 415 and 420 | BNS 318 | Deception, dishonest inducement, delivery of property as the aggravated form | None | The famous 420 survives only as the aggravated limb, Section 318(4) |
| Defamation | IPC 499 and 500 | BNS 356 | An imputation made or published, harming reputation, subject to ten exceptions | None | Community service appears among the punishments for this offence |
| First information report | CrPC 154 | BNSS 173 | Information disclosing a cognizable offence, recorded and read over to the informant | Lalita Kumari | Information sent electronically must be signed within three days |
| Arrest and notice of appearance | CrPC 41 and 41A | BNSS 35 | Necessity as the test for arrest; notice where arrest is not required | Arnesh Kumar | The 41A notice lives inside Section 35, in sub-sections (3) to (6) |
| Remand and default bail | CrPC 167 | BNSS 187 | Investigation periods of sixty and ninety days; bail becomes a right when they expire | None | Police custody may be authorised in parts beyond the first fifteen days |
| Anticipatory bail | CrPC 438 | BNSS 482 | Apprehension of arrest for a non-bailable offence; direction from the Sessions Court or High Court | None | Read this row together with the next one; the pair is the classic trap |
| Inherent powers of the High Court | CrPC 482 | BNSS 528 | Power to quash and to secure the ends of justice | None | Section 482 in the new code is anticipatory bail, not this provision |
| Maintenance | CrPC 125 | BNSS 144 | A summary remedy for wives, children and parents unable to maintain themselves | None | In the old code 144 was the public-order power; the two 144s have swapped codes |
| Urgent orders in cases of nuisance or danger | CrPC 144 | BNSS 163 | A magistrate’s order against apprehended danger or nuisance in urgent cases | None | The provision everyone calls Section 144 is Section 163 in the procedure code in force |
| Confession to a police officer | IEA 25 | BSA 23 | No confession made to a police officer can be proved against an accused | None | Sections 25 to 27 of the old Act are consolidated into the single new provision |
| Statements by a person who cannot be called | IEA 32 | BSA 26 | Statements as to the cause of death are relevant whether or not death was expected | None | The dying declaration is the first limb of a wider provision |
| Opinion of experts | IEA 45 | BSA 39 | Opinion on science, art, foreign law, handwriting or finger impressions is relevant | None | An expert gives material for the court’s decision; the opinion is not proof by itself |
| Admissibility of electronic records | IEA 65B | BSA 63 | A certificate as the condition for relying on an electronic record as secondary evidence | Anvar; Arjun Panditrao | The certificate takes the form in the Schedule, in two parts, with the hash value |
| Burden of proof | IEA 101 | BSA 104 | The party who asserts a fact must prove it | None | The general rule opens the burden chapter in both codes |
| Presumption as to dowry death | IEA 113B | BSA 117 | Where the dowry death ingredients are shown, the court shall presume the death was caused | None | Pair this row with the dowry death row: substance in one code, presumption in another |
Read the lattice the way examiners set questions. One direction is translation: a question names the old provision and asks for its treatment in the new code, and the row answers it. The other direction is recognition: a question gives facts and a new number, and the candidate must identify the doctrine from the elements column before the number means anything. The warning column is where the lattice earns its keep in the last week of revision, because the traps it flags are the ones that appear in papers precisely because paper-setters know candidates will fall into them. The swapped 144s and the 438, 482, 528 triangle are not trivia; they are the cheapest discriminator a paper-setter owns, separating candidates who learned the transition structurally from candidates who learned that the codes had changed and little more.
Two cautions govern the use of any such table, this one included. The first is that a lattice is a revision instrument and not a substitute for the text: where a row shows a provision as consolidated, split or redrawn, the elements column is a memory aid and the bare text governs, and a candidate who suspects a chart of containing an error should test the row against the chapter logic of the code, because a correspondence that places an offence in the wrong neighbourhood of the wrong code announces itself. The second caution is that the lattice records where provisions went, not whether the law changed on the way. Several rows conceal a substantive decision, the marriage age inside the rape definition, the restructuring of police custody inside the remand provision, the disappearance of sedition as a named offence, and a candidate who treats every row as a pure renumbering will answer the comparison questions, which are the questions that reward judgment, at half strength.
Five kinds of correspondence
Not every row of the lattice is the same kind of event, and a candidate who can classify a correspondence has understood it at the level comparison questions test. The first kind is the pure renumbering, where the provision moved and its text travelled with it substantially untouched: the burden of proof rule, the expert opinion provision, most of the definitions of the standard offences. These rows reward nothing but translation, and they are the majority. The second kind is the consolidation, where several provisions were gathered into one: the confession rules of the old evidence code, Sections 25 to 27, live inside the single Section 23 of the Adhiniyam, and the theft definition and its punishment, Sections 378 and 379, live inside Section 303 of the Sanhita. Consolidations are traps for candidates who go looking for the second or third old section as a separate new address. The third kind is the split, where one old provision became two: the cruelty provision, Section 498A, sits at Section 85 in the successor code, with its definitional work done at Section 86, and the old Section 41A notice procedure lives inside Section 35 of the procedure Sanhita as sub-sections (3) to (6) of the arrest provision rather than as a neighbour standing beside it.
The fourth kind is the omission, where the old provision has no successor at all, and omissions must be learned as deliberately as provisions, because a candidate searching a new code for a repealed offence will invent a correspondence rather than report an absence. Adultery is the clean case: Section 497 was struck down in Joseph Shine and the Sanhita simply contains no adultery offence, so the correspondence is a full stop. Section 377 likewise has no successor provision carrying its doctrine; the reading down in Navtej Singh Johar had already confined it, and the new code omitted it. The fifth kind is the fresh provision, with no ancestor: organised crime at Section 111, petty organised crime at Section 112, the terrorist act at Section 113 and snatching at Section 304 of the Sanhita, and community service in the punishment list. Fresh provisions are the only rows that require genuinely new learning of substance rather than of position, and they are correspondingly prominent in papers, because they are the ground on which all candidates start equal. A candidate who tags every lattice row with its kind, renumbering, consolidation, split, omission or fresh provision, owns a study instrument that answers the comparison question’s deepest form, which is not where did this go but what happened to it on the way.
The numbers that function as names
The method has an exception, and the exception is examinable. A small set of section numbers has escaped the statute book and become part of the language: they function as names, used by judges in court, by journalists on front pages, and by examiners in questions, and a candidate who does not recognise them instantly is locked out of the conversation the paper is having. These numbers must be memorised in both systems, not because numbers matter in themselves but because these particular numbers are identifiers, and the test of whether a number belongs to the set is simple: if a judgment, a headline or a senior counsel can use the number without naming the offence and still be understood, the number is a name.
In the substantive code the set is short. Murder is 302 and 103. Rape is 375 and 376, and 63 and 64. Cruelty by a husband or his relatives is 498A, a number so famous it became a noun, and 85 in the Sanhita. Dowry death is 304B and 80. Sedition is 124A, and its successor, the provision on acts endangering the sovereignty, unity and integrity of India, is 152: a candidate must know both that the old number carried the doctrine and that the new provision is a redrawn offence and not the same offence wearing a new label. Cheating is 420 and 318, a number that entered ordinary speech so completely that its survival as Section 318(4), the aggravated limb, is itself a favourite examination point. In the procedure code the names are the bail and information provisions: the first information report at 154 and 173, anticipatory bail at 438 and 482, the inherent powers at 482 and 528, maintenance at 125 and 144, and the public-order power at 144 and 163. In the evidence code the set is smaller still: the confession bar at 25 and 23, and the electronic records certificate at 65B and 63, a number that a generation of lawyers learned as a complete code within the code.
What happened to Section 482 of the Code of Criminal Procedure?
It moved to Section 528 of the Bharatiya Nagarik Suraksha Sanhita, carrying the High Court’s power to quash proceedings and secure the ends of justice. The trap is that Section 482 of the code is a different provision, anticipatory bail, the successor of Section 438, so the old code’s most cited number and its famous neighbour have exchanged places.
The honest way to learn the name-numbers is as pairs with a story attached, because the story is what the examiner is testing underneath the number. The pair 482 and 528 is remembered by the swap. The pair 144 and 163 is remembered by its collision with 125 and 144, the maintenance provision taking the number the public-order provision vacated, so that the two most famous 144s in Indian law have traded codes. The pair 65B and 63 is remembered by the litigation it generated, two Supreme Court decisions and a certificate that became a small industry. A candidate who can tell the story of a number can also answer the follow-up question, which is where the marks are: not what is the new number for anticipatory bail, but why the arrangement of the bail provisions matters, or what the certificate litigation shows about the relationship between a code and the courts that operate it. Numbers memorised as names with stories are also the numbers least likely to be confused under pressure, because each is attached to a doctrine rather than floating in a list.
Everything outside the name set should be learned structurally and looked up without guilt. There are 358 sections in the Sanhita, 531 in the procedure code and 170 in the evidence code, and no examination rewards a candidate for knowing the address of the provision on harbouring an offender or the exact number of the presumption about documents thirty years old. Papers test the architecture, the famous provisions, the doctrines with case law attached, and the changes. A revision plan that spends equal time on all 1,059 sections of the three new codes is not thorough; it is a plan to know everything slightly and the tested things inadequately.
The judgments that carry the marks
Case law in this field is not decoration on the codes; it is the second text, and examiners test it through propositions. The discipline the brief for this guide insists on, and which experienced teachers repeat, is that a proposition stated correctly without a citation scores better than a citation attached to a wrong proposition, because the first error is an omission and the second is a falsehood. The candidate’s unit of learning for a judgment is therefore one sentence: what the court held, stated tightly enough to survive being written from memory. Eight decisions carry a disproportionate share of the marks in this field, and each earns its place by deciding something the codes left open.
A K Gopalan v State of Madras, in 1950, read Article 21 narrowly: the phrase procedure established by law meant a procedure enacted by a competent legislature, and the fundamental rights were treated as separate compartments, so that a detention law was tested against Article 22 and not against the freedoms in Article 19. Maneka Gandhi v Union of India, in 1978, dismantled that architecture: a procedure depriving a person of life or personal liberty must be right, just and fair rather than arbitrary or oppressive, and Articles 14, 19 and 21 interrelate, so that a law must satisfy all three. The pair is the single most productive comparison in the whole syllabus, because almost every later liberty judgment is a footnote to the movement from the first to the second, and a candidate who can state both propositions and the distance between them owns the frame for the detention and arrest material as well.
Lalita Kumari v State of Uttar Pradesh, decided by a Constitution Bench in 2013, held that registration of a first information report is mandatory where the information discloses a cognizable offence, and that a preliminary inquiry is permissible only in a limited set of categories of case. Arnesh Kumar v State of Bihar, in 2014, disciplined arrest from the other end: arrest is not mandatory merely because an offence is cognizable and punishable with imprisonment extending to seven years, the police must record their reasons against the checklist in Section 41 of the old procedure code, and the notice of appearance under Section 41A is the default route where arrest is not needed. Read together, the two decisions are the courts building, on a code that said little, the twin restraints that the syllabus tests constantly: the state must register, and the state must justify taking the person.
What is the one-sentence proposition for which Lalita Kumari is cited?
Where information given to the police discloses a cognizable offence, registration of a first information report is mandatory, not a matter of police discretion, and a preliminary inquiry before registration is confined to limited categories of case, such as matrimonial disputes and commercial offences, and even there it tests only whether a cognizable offence is disclosed at all.
The evidence pair works the same way. Anvar P V v P K Basheer, in 2014, held that Section 65B of the Evidence Act is a complete code for the admissibility of electronic records and that its certificate is mandatory when an electronic record is produced as secondary evidence. Arjun Panditrao Khotkar v Kailash Kushanrao Gorantyal, in 2020, kept the requirement and softened its timing: the certificate is a condition precedent, but it may be produced at a later stage of the proceedings where the party could not obtain it earlier, and the contrary relaxation in Shafhi Mohammad was overruled. The pair matters doubly in a transition syllabus, because the certificate sits at Section 63 of the Bharatiya Sakshya Adhiniyam, in a form set out in the Schedule to that Act, and the examinable question is whether the old case law still governs the new provision, which it does wherever the new text is substantially the same.
The remaining three anchors are the courts amending the substantive code by interpretation. Navtej Singh Johar v Union of India, in 2018, read down Section 377 of the Penal Code so that it no longer criminalised consensual sexual acts between adults, holding the provision unconstitutional to that extent. Independent Thought v Union of India, in 2017, read down Exception 2 to Section 375, raising the age written in the marital exception from fifteen to eighteen so that the provision cohered with the rest of the law on child protection. Joseph Shine v Union of India, in 2018, struck down Section 497, the adultery provision, altogether, as a violation of equality and personal liberty. Each of the three also teaches a technique the examiner can name: reading down to save a provision by narrowing it, in the first two, and striking down where no saving construction was possible, in the third, and a candidate who learns the cases with their techniques attached can answer questions about provisions the cases never touched.
The graded families: where elements earn their marks
The elements method pays its highest dividend inside the graded offence families, where the code arranges a ladder of offences around a single kind of conduct and the examination question asks the candidate to place facts on the right rung. The property family is the classic ladder. Theft is the dishonest taking of movable property out of another’s possession without consent. Extortion adds the element of putting a person in fear of injury and thereby inducing delivery. Robbery is theft or extortion aggravated by violence or the threat of instant violence in its commission. Dacoity is robbery by five or more persons acting together. Cheating runs on a separate rail within the family: deception that dishonestly induces delivery of property or the doing of an act the deceived person would not otherwise do, and criminal breach of trust on another rail again, entrustment followed by dishonest misappropriation, distinguished from cheating by the timing of the dishonesty, present from the start in cheating and arising after a lawful entrustment in breach of trust. Every rung turns on one added element, and problem questions are built by stating facts that sit deliberately between two rungs: property obtained by a threat of future harm sits below robbery and inside extortion; property taken after a scuffle over possession tests whether the violence was used in order to commit the theft. A candidate who has learned the family as a ladder of added elements walks these problems; a candidate who learned each offence as an isolated definition collides with them.
The offences against the body form the second ladder, and its rungs are mental states. Culpable homicide is causing death by an act done with the intention of causing death, or with the intention of causing such bodily injury as is likely to cause death, or with the knowledge that the act is likely to cause death. Murder is culpable homicide carrying one of the aggravated mental states the definition enumerates, subject to the five exceptions, grave and sudden provocation, excess of private defence, excess of a public servant’s lawful power, sudden fight, and consent, which pull a killing back down from murder to culpable homicide not amounting to murder. The punishment provisions track the ladder: Section 105 of the Sanhita for culpable homicide not amounting to murder, Section 103 for murder, Section 106 for the distinct and lesser case of death by a rash or negligent act where the mental state never rises to knowledge. Below death, the hurt ladder runs from simple hurt through grievous hurt, defined by the enumerated kinds of injury, to hurt caused by dangerous weapons and hurt caused in the commission of other offences. The candidate’s discipline here is to argue the mental state from the facts, the weapon, the part of the body, the number of blows, rather than to assert a conclusion, because the ladder is climbed by inference and examiners award the inference, not the label.
The third family a candidate must hold as a unit is the offences against women and children, both because the Sanhita gathers them into a single chapter and because their reform history, through the amendments of 2013 and the judgments reading the old text, is the most examined slice of substantive law in the field. The organising idea of the rape definition is consent, its absence, and the circumstances in which apparent consent is no consent at all: against the woman’s will, without her consent, consent obtained by fear or by impersonation, consent given through unsoundness or intoxication, and consent by a girl below eighteen, which the law treats as no consent regardless of the surrounding facts. Around that centre the chapter arranges the aggravated forms, the custodial and positional offences, the offences of sexual harassment, disrobing, voyeurism and stalking that the 2013 amendments created, and the cruelty and dowry provisions further along the chapter’s run. Learning the family around the consent centre, with the 2013 Act as its modernising moment and Independent Thought as the judgment that completed the age alignment, gives the candidate a chapter that can be reproduced as an argument rather than recited as a catalogue.
The 2013 reform as a set piece
The Criminal Law (Amendment) Act of 2013 deserves study as a set piece, the worked example of how reform happens in this field, because it contains every actor and every mechanism the syllabus elsewhere treats separately. It contains a precipitating event and a committee: the Delhi gang rape and the Justice Verma Committee, constituted within days and reporting within twenty-nine. It contains emergency lawmaking: an ordinance in February 2013 carrying the reform into force before the Act itself was passed, with the Act deemed effective from the ordinance’s date, a sequence whose constitutional mechanics a strong candidate can describe. It contains selective enactment: Parliament took the expanded definition of rape, the new offences of harassment, disrobing, voyeurism and stalking, the raised sentences and the procedural changes on medical examination and on the recording of statements, and declined the committee’s recommendations on marital rape and on the death penalty for rape, so that the statute on the books is a negotiated subset of the report behind it. And it contains the long tail: the death penalty for rape was later introduced for offences against girls below twelve by the amendment of 2018, and the whole body of 2013 law was then consolidated into the Sanhita in 2023, so that a provision a candidate meets at Section 63 of the new code carries a decade of layered history that a good answer can deploy in two sentences.
The set piece also teaches the candidate how to handle a contested legacy without taking a side the question has not asked for. The assessment that the 2013 Act strengthened the law’s response to sexual violence rests on its expanded definitions and new offences; the assessment that it privileged punishment over prevention rests on its sentencing escalations and on the parts of the committee’s institutional programme, policing, prosecution and compensation, that were not enacted with the same energy. Both assessments circulate in the commentary, both are attributed positions rather than facts, and the answer that presents them as such, and then weighs them against the statute’s text, outperforms the answer that adopts either as the writer’s own opening premise. That handling, claim, attribution, text, weighing, is the stance discipline of this whole series compressed into one topic, and the 2013 material is where a candidate should practise it until it is automatic, because the evaluation questions on the 2023 codes will demand exactly the same moves on a larger canvas.
Worked problems: the method under examination conditions
Two extended worked problems show the method handling the genres candidates fear most, the custody confession and the mixed-date prosecution. Take first a problem in which an accused, in police custody, tells the investigating officer where a stolen ornament has been buried, the ornament is recovered from that place, and the prosecution seeks to prove both the statement and the recovery. The candidate’s first move is sorting: this is a proof question, so the evidence code shelf. The second move is the rule: a confession made to a police officer cannot be proved, and a confession made in custody cannot be proved unless made in a magistrate’s immediate presence. The third move is the exception inside the rule: so much of the information as relates distinctly to the fact discovered in consequence of it may be proved, so the part of the statement that led to the ornament’s discovery is admissible and the remainder is not. The fourth move is the transition check: the rule sat at Sections 25 to 27 of the Evidence Act and sits consolidated at Section 23 of the Adhiniyam, the substance unchanged, so the answer can be given once and cited in both systems. Four moves, each from a shelf built earlier in this guide, and the answer assembles itself in the order a court would reason it.
Take a harder problem. A complaint of cruelty and dowry demand is made in May 2024; a first information is registered in June 2024 under the old law; the alleged cruelty continues into August 2024; investigation concludes in October 2024; and the accused seeks anticipatory bail. The candidate must resist the temptation to choose one code for the whole problem. The cruelty alleged before 1 July 2024 is charged under Section 498A of the Penal Code, because the offence was committed under that code; the conduct alleged after that date falls under Sections 85 and 86 of the Sanhita, and whether a continuing course of conduct can be charged as a single matter across the line is an argument to be made from the savings provisions and the nature of the offence, stated as an argument rather than assumed. The investigation and its steps are governed by the procedural savings: proceedings pending at commencement continue under the old procedure code as if the new one had not come into force, which is what Section 531 of the procedure Sanhita provides, and the bail application is worked within that architecture, with the candidate noting that anticipatory bail in the new code sits at Section 482 and in the old at Section 438, and citing the one that matches the proceeding. The evidence, when the matter reaches trial, engages the Adhiniyam’s repeal and savings in Section 170 and the continuity of doctrine where provisions correspond. A problem of this shape is not five questions; it is one question asked five ways, and the candidate who answers it code by code and date by date, saying so in the answer’s structure, demonstrates the exact competence the transition papers exist to certify.
The third worked exercise is of a different kind: not a problem question but a fault-finding drill, the genre in which a paper presents a short passage containing planted errors. A passage describing a 2025 prosecution might state that the accused was charged with cheating under Section 420 of the Bharatiya Nyaya Sanhita, that his statement to the police was proved against him under Section 23 of the Code of Criminal Procedure, that the High Court quashed the proceedings under Section 482 of the Sanhita, and that the magistrate imposed a prohibitory order under Section 144 of the new procedure code. Each sentence carries its error. Cheating is Section 318 of the Sanhita, 420 being the Penal Code number, and only the aggravated limb at that. The confession bar is evidence law, Section 23 of the Bharatiya Sakshya Adhiniyam, not procedure at all. The inherent power to quash is Section 528 of the procedure Sanhita; Section 482 of that code is anticipatory bail, and the Sanhita is the substantive code in any event, which contains no quashing power. The prohibitory order power is Section 163 of the procedure Sanhita; Section 144 there is maintenance. Four errors, four traps from the checklist, and a candidate who has drilled the checklist reads the passage the way a proofreader reads, which is precisely the reading the fault-finding genre rewards. Building such passages for oneself, from the trap list, is among the most efficient revision exercises available in the transition years, because it trains detection rather than recall, and detection is the scarcer skill.
The boundaries of the general codes
The general codes do not exhaust the criminal law syllabus, and a candidate should know where their territory ends, because questions at the boundary are set deliberately. Special statutes occupy ground the general codes leave to them: the preventive detention statutes and the special security legislation operate beside the Penal Code rather than inside it, with their own definitions, their own procedure and their own constitutional testing ground in Articles 21 and 22. The study discipline at the boundary is the same three-question sort, applied to a different shelf: is the question about the special statute’s own machinery, about the general code offence that parallels it, or about the constitutional frame that governs both? A candidate who cannot say which is being asked will answer a detention question with arrest doctrine and an arrest question with detention doctrine, and the two are tested precisely because they are neighbours with opposite rules, arrest being punitive process following an offence and detention being preventive action taken before one. The detailed treatment of the detention and security cluster belongs to the other guides in this series, and this guide’s contribution to the boundary is the habit: sort first, and let the sort choose the shelf.
Procedure inside the general code also changed in texture in ways the candidate should be able to instance beyond the renumbering, because the difference between a renumbering and a change is the master distinction of the whole syllabus. The procedure Sanhita writes timelines into the process in a way the old code largely did not: the judgment in a trial is to be pronounced within thirty days of the conclusion of argument, a period the code permits to be extended to forty-five days for reasons recorded, and the investigation of the specified offences against women and children carries a direction that it be completed within two months of the information being recorded. Some of these timelines are new in kind and some restate earlier directions with statutory force; the candidate’s task is to instance them accurately rather than to generalise about speed. The evidence code’s texture changed at the certificate, where the Schedule prescribes the form, and the substantive code’s texture changed in its punishments, community service being introduced as a punishment in its own right for a small set of offences including the aggravated form of defamation and petty theft by a first-time offender who restores the property. Texture examples of this kind, few, accurate and provision-specific, are what separate an answer about the reform from an answer about the idea of the reform.
The transitional rule and how it is tested
One rule of the transition is itself examinable, and candidates get it wrong with impressive consistency. The three new codes came into force on 1 July 2024, and the line they draw is the date of the offence for substance and the pendency of the proceeding for procedure. An offence committed before that date is governed by the Indian Penal Code, because Article 20(1) of the Constitution forbids conviction under a law that was not in force when the act was done and forbids a penalty greater than the one the law then provided, and Section 358 of the Bharatiya Nyaya Sanhita, the repeal and savings provision, preserves liability incurred under the repealed code. Proceedings pending when the Bharatiya Nagarik Suraksha Sanhita commenced continue under the old procedure code, which Section 531 of the new code says in terms, providing that pending investigations, inquiries, trials and appeals are to be disposed of under the Code of Criminal Procedure as if the Sanhita had not come into force. The Bharatiya Sakshya Adhiniyam carries the same architecture in its Section 170. An offence committed on or after 1 July 2024 belongs to the new codes from the first information onwards.
The reason examiners love this rule is that it can be tested in both directions and at three depths. The shallow test is a date: an offence committed in June 2024, a first information registered in August 2024, and the candidate must resist the pull of the registration date and answer with the offence date for substance. The middle test is a stage: a trial pending at commencement, and the candidate must say which procedure code carries it to judgment and which code carries the appeal. The deep test is a doctrine: why the substantive line is constitutional and not merely statutory, which requires Article 20(1) and the ex post facto principle, and why the procedural line is a legislative choice rather than a constitutional command, procedure being generally a matter of the law in force when the step is taken unless the legislature saves the old regime, which here it largely did. A candidate who can work all three depths owns the transition; a candidate who memorised the commencement date owns one fact. This discussion is general legal information offered for study; anyone facing an actual proceeding should have the governing code for their matter confirmed against its dates and record, since the savings provisions reward precision and punish assumption.
One refinement separates strong answers from adequate ones. The savings provisions do not freeze the old codes in every respect: anything done under the repealed code is deemed to have been done under the corresponding provision of the new one, and the courts have the ordinary work of deciding, provision by provision, whether an old-code step survives into a new-code stage. A candidate should therefore present the rule as a strong default with a statutory texture, not as two sealed worlds. That is also the honest position, and examiners reward the candidate who states the rule, states its source, and states its texture over the candidate who states the rule as a slogan.
The confusions that cost marks
Certain confusions recur so reliably that a candidate can treat them as a checklist of things to have unlearned. The first is the names of the three new codes themselves. The Bharatiya Nyaya Sanhita is the substantive code, the offences. The Bharatiya Nagarik Suraksha Sanhita is the procedure code, and the word Suraksha, protection, is the hook: it protects the citizen inside the process. The Bharatiya Sakshya Adhiniyam is the evidence code, and Sakshya means evidence or witness testimony. A candidate who attaches each name to its function once, in those terms, stops confusing them; a candidate who learns them as three similar strings will confuse them in the examination hall, and the error is costly because it poisons every answer that follows. The detailed treatments in the Code of Criminal Procedure guide and the Bharatiya Nagarik Suraksha Sanhita guide show the procedural pair side by side, and the same pairing discipline applies to the Indian Evidence Act guide and the Bharatiya Sakshya Adhiniyam guide.
The second confusion is answering across the codes: giving the punishment for an offence when the question asked for the stage of trial, or the definition of a confession when the question asked when it may be proved. The third is citing the number without the code, an error the transition has made dangerous, because Section 482, Section 144 and Section 63 each mean different things depending on the code attached, and an answer that writes a bare number in a transition paper is an answer that may be right in one code and wrong in the answer’s own terms. The fourth is assuming that a changed number means changed law, or that an unchanged name means unchanged law: the remand provision kept its function and altered the custody rules inside it, while the sedition provision changed both its number and its content and lost its name in the process. The fifth, and the most quietly expensive, is writing about the new codes as though the old ones had vanished from the legal system, when every pending trial in the country is a standing refutation: the old codes govern the largest body of live criminal litigation there is, and an answer that cannot say so has misunderstood the subject it is describing.
What carries over and what must genuinely be relearned
The anxiety that dominates discussion among aspirants and law students is that everything previously learned has been wasted. It has not, and a candidate is better served by an exact account of the inheritance than by reassurance. The elements of the offences carry over: murder, theft, cheating, defamation and the rest are defined in the successor code in terms that track the old definitions closely, with consolidation rather than reinvention as the dominant technique. The general exceptions carry over, the architecture of abetment and common intention carries over, the procedural sequence from information to appeal carries over, and the evidentiary logic of relevancy, proof, burden and presumption carries over almost intact. The case law carries over wherever the successor provision is substantially the same as the provision it replaced, which is most of the litigated ground: Anvar and Arjun Panditrao govern the certificate at Section 63 of the evidence code because Section 63 is Section 65B re-enacted with a prescribed form, and the arrest jurisprudence governs the notice provisions of the new procedure code because those provisions codified what the courts had built.
What must genuinely be relearned is narrower and should be listed without flinching. The numbering must be relearned, at least for the name provisions and the lattice rows. A defined set of substantive changes must be learned as changes: the new offences of organised crime at Section 111, petty organised crime at Section 112 and the terrorist act at Section 113 of the Sanhita; snatching as a free-standing offence; community service as a punishment; the removal of sedition as a named offence and its replacement by the differently drawn Section 152; the disappearance of adultery as an offence, and the narrowing of attempted suicide to the single case of an attempt meant to compel or restrain a public servant in the discharge of duty; the restructuring of police custody within the remand provision; and the electronic information and timeline provisions of the procedure code. That list is finite, and finitude is the point. A candidate who treats the change as total will waste the transition year relearning theft; a candidate who treats it as cosmetic will be caught by every item on the list. The syllabus rewards the candidate who can say, provision by provision, which of the two it is. The Article 22 and preventive detention side of the syllabus, where the transition bites least, is owned by the civil liberties examination guide, and a candidate should take that cluster from there rather than from this one.
The trap list, completed
The confusion checklist given earlier in this guide names the famous traps; a completed list adds the traps that sit one level down, which papers use to separate strong candidates from merely prepared ones. The first of these is the collision of Section 304. In the Penal Code, Section 304 is the punishment for culpable homicide not amounting to murder. In the Sanhita, Section 304 is snatching, a property offence, and the culpable homicide punishment has moved to Section 105. A bare citation of 304 therefore reverses its meaning between the codes, describing a killing in one and a chain pulled from a neck in the other, and a candidate’s answer that does not attach the code to the number is a coin toss. The second is the hurt pair: simple hurt at Section 323 of the old code is Section 115 of the new, and grievous hurt at Section 325 is Section 117, while the numbers 323 and 325 in the new code belong to entirely different territory; the pair is dangerous because the old numbers are drilled into every candidate’s reflexes from the first year of study. The third is the pair of presumptions about abetment of suicide and dowry death, Sections 113A and 113B of the Evidence Act, which sit at Sections 116 and 117 of the Adhiniyam, close enough to their old addresses to feel unchanged and different enough to mark an answer as careless, and which must in any event be cited with their substantive partners, the dowry death provision in the substantive code, because the presumption and the offence are tried together and examined together.
The fourth trap is one of names within the substantive code. Section 63 of the Sanhita is the definition of rape, and Section 63 of the Adhiniyam is the electronic records provision, and Section 63 of the Penal Code does not exist as a famous provision at all; the number therefore tests whether a candidate attaches codes by reflex, since a bare 63 in an answer can be right about evidence and wrong about substance or the reverse. The fifth trap concerns kidnapping and abduction: kidnapping from lawful guardianship sat at Section 363 of the Penal Code and sits at Section 137 of the Sanhita, and the definitions around it, kidnapping from India, abduction, and the aggravated forms for ransom and for compelling marriage, moved as a family whose internal relations matter more than any single address. The sixth trap is a trap of language rather than number: candidates describe Section 152 of the Sanhita as sedition, because the commentary does, and thereby assert the very thing the comparison question is testing, that the successor provision reproduces the old offence; the provision punishes acts endangering the sovereignty, unity and integrity of India, it does not use the word sedition, and its elements differ from Section 124A’s disaffection formula, so the accurate answer names it by its own words and then conducts the comparison. The seventh trap is the mirror of the sixth: describing community service as an old punishment newly named, when in the substantive code it is a genuinely new entry in the punishment list, or describing the removal of the attempt to commit suicide as a complete decriminalisation, when the Sanhita retains the narrow offence of attempting suicide in order to compel or restrain a public servant in the discharge of duty.
A candidate should convert this list, and the famous traps from the earlier checklist, into a personal register that grows with every practice paper. The register’s format matters less than its rule: every error in practice is entered with the trap that produced it and the correction stated as a correspondence, never as a bare fact. Over a season the register becomes the most individual document the candidate owns, a map of their own failure modes rather than of the syllabus, and the final week’s revision should work from the register first and the lattice second, because the lattice describes the system while the register describes the candidate, and marks are lost at the intersection of the two.
Answer architecture for descriptive papers
A descriptive answer in this field has a load-bearing order, and candidates who follow it find that even answers they are unsure of earn their structure marks. The answer opens by identifying the issue in one or two sentences, naming the code and the doctrine the question engages, so that the examiner knows within three lines which shelf the answer is coming from. It then states the governing provision with its code attached and its elements or conditions in sequence. It then deploys the proposition from the governing judgment where one exists, stated in a sentence, attributed, and bounded. It then applies the law to the facts the question actually gave, taking the facts in the order of the elements, and dealing with the date and the code choice expressly wherever the facts cross the transition line. It closes with a conclusion that answers the question asked, not the question the candidate would have preferred, and where the question invites evaluation, it closes with the differentiated verdict the evaluation section of this guide models. Five movements: issue, provision, proposition, application, conclusion. The order is not a formula for its own sake; it mirrors how a court reasons, which is why examiners who are or were practitioners reward it without thinking about why.
Two refinements distinguish the top band. The first is the express treatment of the counter-position at the point where it bites: in an evaluation answer, the rival assessment stated at its strongest before the weighing; in a problem answer, the rival code or the rival rung of the ladder acknowledged and excluded by an element, so that the examiner sees the candidate has considered the answer they did not give. The second is economy with authority: one governing provision and one governing proposition, fully used, outscore five of each name-checked, and the candidate who cites the Constitution Bench decision that governs and explains its boundary has done more than the candidate who decorates the answer with every case in the chapter. Both refinements are habits, and both are trainable in practice answers long before they are needed in the hall.
The evaluation questions: stating the reform debate honestly
The evaluation shape of question asks the candidate to judge the reform itself, and it is the shape where the discipline of attribution matters most, because evaluation questions invite slogans. The government’s case for the 2023 codes, as its proponents state it, rests on four claims: that the colonial codes were instruments of subjugation designed for ruling subjects rather than serving citizens, and that their replacement is an act of legislative self-respect; that the new codes reorder priorities towards victims, placing offences against women and children at the front of the substantive code; that timelines written into the procedure code, for investigation, for committal, for judgment, will discipline a system notorious for delay; and that the new scheme embraces electronic process, from electronic information to electronic records, as the colonial texts could not. The critics’ case, as its proponents state it, answers in kind: that a change of names and arrangement is not decolonisation where the substantive architecture of police power survives intact and in places expands, as in the restructured custody provisions; that the renaming was carried through Parliament at a speed and in circumstances that limited scrutiny; that widening provisions on organised crime and terrorism-adjacent offences in a general code risks importing exceptional law into ordinary law; and that several changes praised as modernising, including the removal of sedition’s name, conceal continuities or broader formulations, Section 152 being the standard exhibit.
A candidate earns marks in this debate by doing three things. The first is stating each side at its strongest and in its own terms, so that the government’s decolonisation claim is tested against the statutory record rather than against a caricature, and the critics’ continuity claim is tested provision by provision rather than assumed. The second is bringing the text to the argument: the placement of the women and children chapter, the content of the custody provisions, the enumerated timelines, the fate of specific offences, are evidence, and an evaluation answer without provision-level evidence is opinion wearing a gown. The third is reaching a verdict where the record supports one, and saying precisely where it does not: the record supports the conclusion that the reordering and consolidation are real, it supports the conclusion that the police powers architecture was re-enacted rather than rethought, and it does not support either the claim that nothing changed or the claim that everything did. An answer that lands on that kind of differentiated conclusion has done what the question asked; an answer that picks a side in the first paragraph and defends it for five pages has answered a different and easier question.
Committees supply the honest candidate with a paper trail that predates the slogans. The Justice Verma Committee of 2012 and 2013 demonstrated what a focussed reform exercise looks like, receiving a defined mandate after the Delhi gang rape, reporting in twenty-nine days, and seeing a substantial part, though never the whole, of its programme enacted within weeks, including the parts Parliament declined, on marital rape and on the death penalty for rape, which remain part of the record candidates should be able to state. The Committee on Reforms of the Criminal Justice System under Justice Malimath, reporting in 2003, supplies the longer view, having proposed systemic changes to investigation, prosecution and victim compensation two decades before the new codes, some of which the codes reflect and many of which they do not. The Law Commission’s forty-first report underlies the 1973 procedure code. A candidate who can place a reform claim against the committee record, accepting what the record shows and noting what was recommended and not taken, writes evaluation answers of a different order from the candidate who has only the press release and the counter-slogan.
Two papers, two candidates: the civil services and the judiciary
The same three codes feed two examination cultures, and a candidate should know which culture is being fed. The civil services candidate meets criminal law inside a wider polity and governance paper: the demand is structural understanding, the ability to place the codes inside the constitutional frame of Articles 20, 21 and 22, to evaluate reform as public policy, and to explain the transition’s significance to a lay reader in clear prose. Section-level precision matters less than architecture and judgment, though the name provisions and the leading cases are still the currency in which answers are cashed. The judiciary candidate meets the codes as working instruments: the demand is provision-level accuracy, the staging of a trial, the drafting sensibility that notices a missing certificate or a wrong forum, and the ability to apply a provision to facts in the sequence a court would follow. For that candidate the lattice is not a revision aid but a map of the courtroom, and the fault-finding question is the home genre.
The practical consequence is a difference of emphasis within a shared foundation. Both candidates need the three shelves, the four moments, the elements method and the propositions. The civil services candidate should weight the evaluation material, the committee record and the constitutional frame; the judiciary candidate should weight staging, bail, the confession and certificate rules, and the transitional rule at its middle depth, where a pending proceeding has to be assigned to its code stage by stage. A candidate preparing for both, a common position, should build the shared foundation once and then run two specialised passes, rather than treating either paper as a diluted version of the other. Coaching material that ignores the distinction serves neither candidate, because it teaches policy candidates to memorise numbers they will never be asked for and teaches judiciary candidates to gesture at reforms they are expected to operate.
A word on sources closes this part of the guide, because transition study multiplies the candidate’s exposure to error. The hierarchy of trust is simple to state and worth obeying: the bare text of the codes, as enacted and as amended, stands at the top; the committee and commission reports stand next, for the history and the stated reasons; reported judgments stand next, read for their propositions in the official reports where they can be got; and commercial charts, summaries and digests stand at the bottom, useful as indexes and dangerous as authorities. Every contested correspondence in the lattice in this guide has been checked against the structure of the codes themselves, and the anchor test set out in this guide, testing any chart against the murder provision and the certificate provision before trusting its other rows, exists precisely so that a candidate can repeat the check on any source, including this one. A candidate who works from the top of the hierarchy downwards will find that the bottom of it becomes largely unnecessary, which is the final economy the elements method buys.
Reading the syllabus and the material market honestly
A transition multiplies the volume of advice in circulation, and a candidate needs a filter for it as much as for mapping charts. The primary document is the syllabus notification of the examination actually being attempted: it fixes the papers, the subjects and, in a transition, often the treatment of the codes themselves, and no coaching digest outranks it. Where a notification prescribes the new codes, the old codes remain worth their place for the reasons this guide has given, litigation reality and the reading of judgments, but their weight in the revision plan follows the notification and not habit. Where a notification is ambiguous, the candidate’s protection is the shared foundation: the elements, the architecture and the propositions serve under either prescription, and only the numbering emphasis shifts, which is a week’s adjustment rather than a season’s loss. The same honesty applies to claims made for material in the market. A chart is worth its anchors, a digest is worth its fidelity to the bare text, and a claim that a paper will certainly be set on one code or the other, made by anyone other than the examining body, is a marketing statement and should be priced as one.
The candidate’s own records deserve the same unsentimental treatment. A preparation log that records hours is a weak instrument; a record that logs errors against traps, propositions that failed cold recall, and lattice rows that reversed under pressure is a strong one, because it measures the things the examination measures. The strongest single predictor available to a candidate in this field is the cold production test: given a doctrine, can the candidate produce the provision in both codes, its elements, its governing proposition and its trap, without looking? Everything in a revision plan should be organised to increase the number of doctrines that pass that test, and anything in the plan that does not contribute to it, however industrious it feels, is displacement. This is also the answer to the perennial question of how many hours the codes deserve: they deserve the hours the test says they deserve, measured by production and not by exposure, and the measurement should happen weekly so that the plan can be corrected while correction is still cheap.
There is a final navigational point about the languages of the codes themselves. The new codes carry Hindi names and Sanskrit-derived titles, and examination questions in some papers reproduce those names in full; a candidate should be able to recognise each code by its full name, by its short name and by its function interchangeably, since a question that names the Bharatiya Nagarik Suraksha Sanhita in full is asking about procedure however unfamiliar the string looks under pressure. The names are also, handled properly, a study aid rather than an obstacle: Nyaya for the code that defines wrongs and their just deserts, Suraksha for the code that protects the person inside the process, Sakshya for the code of proof and testimony. Anchoring each to its meaning once, as this guide did at the start, converts three confusable strings into three labels that explain themselves, and the conversion is one of the cheapest gains available anywhere in the transition syllabus.
What complete preparation looks like
Complete preparation in this field can be described as a small number of possessions. The candidate possesses the three shelves and can sort any provision or question onto them without hesitation. The candidate possesses the four moments and can narrate the field’s history as an argument about codification, separation, reform and replacement rather than as a chronology. The candidate possesses the lattice in both directions for the provisions that dominate papers, and the name numbers as pairs with their stories. The candidate possesses one-sentence propositions for the governing judgments, knows their weight and their boundaries, and has the candour discipline for the cases whose citations are not certain. The candidate possesses the transitional rule at three depths and the trap register built from their own errors. And the candidate possesses an answer architecture practised enough to be invisible. None of these possessions is a fact; all of them are capacities, which is why they survive a renumbering that destroyed so many lists, and why the candidate who has spent the transition building them will find the next reform, whenever a legislature produces one, a smaller event than this one.
Drills that build the method
The method in this guide is trainable through five drills, and a candidate’s revision week should rotate them rather than repeat whichever feels most comfortable. The translation drill takes a page of the lattice, covers one column, and produces the missing correspondence together with the elements, against a timer; it builds the recognition speed that objective papers measure. The proposition drill takes the judgment list, covers the holdings, and states each proposition from the case name alone, marking every proposition that cannot be produced cold for rewriting into the personal proposition file; it builds the written paper’s core currency. The chapter walk takes a neighbourhood of the substantive code and requires the candidate to narrate it, the family of offences, their graded relations, and the judgment attached to each, without notes; it builds the finding system that problem questions reward. The staged problem drill takes a single set of facts and works it through information, arrest, remand, charge and trial in order, choosing the code at each stage and saying why; it builds the procedural sequence as a habit rather than a chart. The trap passage drill, described in the worked problems section of this guide, has the candidate compose a short passage containing four planted errors from the trap register and exchange it with a study partner or correct it after an interval; it builds detection, the skill the fault-finding genre exists to measure.
What the drills share, and what distinguishes them from rereading, is production under mild pressure with immediate checking. Rereading produces familiarity, which feels like knowledge and behaves like knowledge until the moment it is asked to perform; production exposes the gap while there is still time to close it. A candidate in the transition years faces a specific temptation, to treat the codes as news, something to be kept up with by reading summaries, and the drills are the corrective, because the examination will not ask what the candidate has read about the reform but what the candidate can do with it: sort it, translate it, apply it, detect its errors and judge it. Five drills, rotated, measured by the cold production test, are a complete training system for the field as this guide has mapped it, and they share one further virtue: they work unchanged for any future renumbering, because they train the structure and not the season.
Why both codes remain working law
One study error the transition invites deserves direct correction, because it is made by able candidates: the assumption that the old codes are an archive, of interest to historians and examiners but dead in the courts. The opposite is true, and the reason is arithmetic. Every offence committed before 1 July 2024 is a Penal Code matter in substance, and the volume of such matters is the accumulated backlog of the criminal courts together with every offence committed in the years up to the change that has yet to be investigated, charged or tried. Investigations opened under the old procedure continue under it; trials pending at commencement run to judgment under the code they began with; appeals from those judgments carry the old codes with them. A junior lawyer joining a criminal practice will therefore spend the working week inside the repealed codes, citing Section 302 and Section 482 in their old senses to courts that use them without any sense of anachronism, and will meet the new codes at the first information desk rather than in the trial court. The judiciary candidate who will join that world is examined on both systems because the job is both systems, and the civil services candidate is examined on the transition because administering it, the circulars, the training, the reprinting of registers and the retraining of investigating officers, is a governance task of exactly the kind the service exists to perform. Study follows function here as everywhere in this guide: the codes a candidate will use are the codes the candidate should be able to work in, and for the span of a career beginning in the transition years, both sets are working law.
The practical consequence for the revision plan is that old-code study is not a concession to examiners but a professional investment with a long yield. The candidate should therefore resist every plan that quarantines the old codes into a single comparative chapter at the end of a book or a course. The correspondences should be carried inside every topic as it is learned, offence by offence and stage by stage, so that the two systems are one knowledge with two vocabularies. That arrangement costs a little effort at the point of learning and saves a great deal at the point of use, in the hall and afterwards, which is the exchange this entire guide has been recommending in one form or another from its first page. The habit has a further return in the examination hall itself, where papers set during the transition years will continue to quote the old codes in questions, in options and in the extracts printed for comment, and the candidate who has kept both vocabularies warm reads those papers at full speed while the candidate who discarded the old law for the new spends the first minutes of every such question on translation rather than on thought.
How questions are framed, and how to work them
Questions in this field are framed in a small number of recurring shapes, and recognising the shape is half the answer. The translation shape gives a provision in one code and asks for its position in the other, and it rewards the lattice learned in both directions. The proposition shape names a judgment, or describes a fact situation a judgment decided, and asks what the law is; it rewards the one-sentence proposition and punishes the candidate who can only narrate the case. The problem shape gives facts, usually with a date buried in them, and asks which code governs and what follows; it rewards the transitional rule and the elements method, since the facts must be sorted into elements before any provision applies. The evaluation shape asks whether the reform achieved its stated aims, decolonisation, speed, victim protection, and it rewards the stance discipline of this series: the government’s case and the critics’ case each at their strongest, attributed to the people who actually make them, and a verdict where the record supports one. The fault-finding shape, beloved of judiciary papers, plants an error, a wrong code, a wrong forum, a missing certificate, and asks the candidate to find it; it rewards exactly the confusion checklist set out in the previous section.
The worked approach to a typical problem question shows the method in sequence. Take a problem in which a man is alleged to have taken a phone from a shop counter in May 2024, the first information is registered in August 2024, and the prosecution tenders a printout of camera footage without a certificate. The candidate who works by elements begins by dating the offence: May 2024 places the substance under the Penal Code, so the charge is theft under the old code, and the taking, the dishonesty and the absence of consent are the elements to address, with snatching unavailable because the new offence did not exist when the act was done. The proceeding, pending at the commencement of the new procedure code or begun under the savings architecture, is then placed under the procedural savings rule rather than assumed into the new code. The evidence question is answered from the certificate jurisprudence: the printout is secondary electronic evidence, the certificate is a condition of relying on it, Anvar states the rule and Arjun Panditrao its timing, and the corresponding provision in the new evidence code is identified in case the stage is governed by it. Three codes, three questions, each answered from its own shelf: that ordering, visible on the page, is what a structured answer looks like, and it is achievable by any candidate who has built the shelves.
Prelims and Mains divide the same material by demand rather than by content. A Prelims question asks for recognition under time pressure: which provision, which case, which code, and it is defeated by the confusion traps and defeated only by them, since the underlying knowledge is usually present. A Mains or judiciary answer asks for construction: elements stated, provisions cited with their codes attached, propositions deployed, a conclusion reached. The revision consequence is that the same lattice serves both, read quickly for recognition in the objective season and written out in prose, row by row, for the descriptive season, because writing a row out as a paragraph is the exercise that converts recognition into construction.
A revision structure that survives the transition
The revision structure follows from everything before it, and it has four passes. The first pass builds the shelves: the three codes and their functions, the four moments, the chapter architecture of the substantive code, the procedural sequence, the evidentiary logic. The second pass stocks the shelves: the lattice, row by row, with the elements column learned as prose rather than as table entries. The third pass attaches the propositions: one sentence per judgment, written out by hand, in a personal proposition file that the candidate builds and keeps across attempts, because a transition syllabus punishes the candidate whose notes belong to a single season. Keeping that file well is a discipline in itself; the habit of maintaining a private, structured research file, protected and portable, is the same habit described in how journalists use VaultBook to protect sources, and a candidate’s proposition file deserves the same care as a reporter’s notebook, since both are useless if lost and dangerous if muddled. The fourth pass is fault-finding against oneself: previous questions and standard material worked with the confusion checklist open, every error logged against the trap that produced it, until the traps stop firing.
Two habits protect the whole structure. The first is to cite, in every practice answer, the code with the number, never the number alone, until the habit is automatic; it costs seconds and it removes an entire class of error. The second is to sanity-check any mapping chart, including the lattice in this guide, against two anchors whose positions are certain, the murder provision and the certificate provision, before trusting its other rows, because a chart that is wrong on an anchor is wrong somewhere else as well, and the published charts in circulation contain errors that this simple test exposes.
The verdict for candidates
The transition between the codes is a change of address, not a change of country. The candidate who learns it as a change of country will spend the season in grief for a syllabus that has not actually been lost; the candidate who learns it as a change of address, keeping the elements, the architecture and the propositions and re-mapping the numbers that function as names, will find that the new codes are the old subject in a better order, and that the questions examiners can ask about the change are among the most predictable on the paper. Elements before numbers. That is the whole method, and it survives any renumbering a legislature can devise.
Frequently Asked Questions
Q: How should aspirants study the shift to the new criminal laws?
Study the shift as a change of address rather than a change of subject. Build the three shelves first: substance, procedure and proof, and learn which code, old or new, answers which kind of question. Then work through the correspondences by offence and doctrine, holding each provision by its elements and its chapter position, so that a number in either system can be traced to the doctrine it names. Learn the small set of numbers that function as names in both systems, attach one-sentence propositions to the leading judgments, and treat the transitional rule, the offence date for substance and pendency for procedure, as a topic in its own right, because it is examined as one.
Q: Do judiciary exams still test the IPC and the Code of Criminal Procedure?
The honest answer depends on the examining body and the syllabus notification for the examination in question, and a candidate should read that notification rather than rely on general advice. The structural point survives every notification: the old codes govern every offence committed before 1 July 2024 and every proceeding saved under the transitional provisions, so the courts will be applying the Penal Code and the old procedure code for years, and a judicial officer who cannot work in them cannot work at all. Examinations that test the working law therefore have every reason to test both systems, and the candidate who abandons the old codes on the strength of a rumour is taking a risk no notification has justified.
Q: Which section numbers are worth memorising in the new codes?
Memorise the numbers that function as names: the provisions that judges, journalists and examiners use as identifiers without naming the offence. In the Bharatiya Nyaya Sanhita that means murder at 103, rape at 63 and 64, cruelty at 85, dowry death at 80, the sovereignty provision at 152 and cheating at 318. In the procedure code it means the first information report at 173, anticipatory bail at 482, the inherent powers at 528, maintenance at 144 and the urgent public-order power at 163. In the evidence code it means the confession bar at 23 and the electronic records certificate at 63. Everything else should be retrievable from elements and chapter position rather than carried as a bare number.
Q: How do you avoid mixing up the three new criminal laws?
Attach each name to its function and refuse to learn them as three similar strings. The Bharatiya Nyaya Sanhita is the substantive code: it defines offences and punishments, and Nyaya, justice, is the hook. The Bharatiya Nagarik Suraksha Sanhita is the procedure code: Suraksha, protection, is the protection the process owes the citizen, from first information to appeal. The Bharatiya Sakshya Adhiniyam is the evidence code: Sakshya means evidence and testimony, and the Act governs proof. A second safeguard is mechanical: never write a section number in an answer without the code attached to it, because most mix-ups become visible the moment the writer has to name the statute the number is claimed from.
Q: Which criminal law judgments are most frequently tested?
The judgments that carry the marks are the ones that decided what the codes left open. A K Gopalan and Maneka Gandhi frame Article 21, the narrow reading of 1950 against the just, fair and reasonable procedure of 1978. Lalita Kumari makes registration of a first information report mandatory where a cognizable offence is disclosed, and Arnesh Kumar restrains arrest where it is not necessary. Anvar and Arjun Panditrao govern the electronic records certificate, first as a complete code and then on the timing of the certificate. Navtej Johar read down Section 377, Independent Thought read the marital exception age up to eighteen, and Joseph Shine struck down adultery. Learn each as one proposition, not as a story.
Q: How should you answer a question comparing old and new codes?
Build the answer on correspondences, not on parallel summaries. Open with what the two provisions share, the elements that carried over, because most of the substantive law did. Then isolate the change precisely: a renumbering, a consolidation, a redrawing, or a genuine change of substance, and name which of the four it is, since comparison questions reward the classification. Then give the consequence, for a court, for an accused, or for a candidate citing the provision, and close with a judgment where the record supports one. Avoid the two failure modes: describing each code in turn and leaving the comparison to the reader, and treating every renumbering as either nothing or everything.
Q: Is it worth learning IPC section numbers now?
Yes, for two distinct reasons, and the reasons dictate how much weight to give them. The first is litigation reality: the Penal Code governs every offence committed before 1 July 2024, so its numbers remain the working vocabulary of pending trials, of the judgments candidates must read, and of most standard commentary. The second is examination design: papers in a transition period cite either code, and a candidate who knows only the new numbers cannot read a question framed on the old ones. The qualification is method. Learn the old numbers as one half of a pair, attached to the elements and the new correspondence, and learn the famous ones as names; do not invest in the old list as a free-standing memory load.
Q: What is the commonest error in criminal law reform answers?
The commonest error is answering from the wrong code while sounding entirely confident: citing the punishment from the substantive code when the question asked about proof, quoting a section number without naming its statute in a field where 482, 144 and 63 each mean different things in different codes, or applying the new code to an offence committed before it commenced. A close relative is treating renumbering as either pure cosmetics or total transformation, when the record is provision-specific. The defence is procedural. Sort the question into substance, procedure or proof before writing, date the facts before choosing the code, and classify each correspondence before describing it.
Q: How should a candidate revise the criminal codes in the final month before an examination?
Spend the month on retrieval, not on fresh reading. Work the correspondence lattice in both directions until translation is automatic, write the one-sentence propositions out from memory and check them against the file, and drill the traps deliberately: the swapped public-order and maintenance numbers, the anticipatory bail and inherent powers pair, the certificate provision, and the transitional rule worked at its three depths. Then write full answers against the clock, citing the code with every number, and log every error against the trap that produced it. New material in the final month has a poor exchange rate; the marks available in this field sit in accuracy under pressure, which is a trained skill rather than a stock of facts.
Q: How do Prelims and Mains questions on criminal law differ in what they demand?
A Prelims question demands recognition under time pressure: the right provision, the right case, the right code, chosen from options designed around the known confusions. It is defeated less by ignorance than by the traps, the swapped numbers and the plausible wrong code. A Mains or judiciary answer demands construction: elements stated, provisions cited with their statutes named, propositions deployed in sequence, and a conclusion reached and defended. The preparation differs accordingly. Objective revision is speed work on the lattice and the name numbers; descriptive revision is writing rows of the lattice out as prose paragraphs, because the movement from recognising a correspondence to constructing an answer from it is a separate skill that only writing trains.
Q: How do you learn an offence by its elements rather than by its section number?
Take the offence apart into the four questions every definition answers. What conduct is required, an act, an omission, a possession, a publication? In what circumstance must it occur, against whom, where, with what relationship between the parties? What result, if any, must follow? And with what state of mind must the act be done, intention, knowledge, dishonesty, rashness or negligence? Write the four answers in a line each and the offence is learned in a form that survives renumbering, because the successor code defines the same offences through the same four questions. The number then attaches to the structure as a label, in either code, instead of floating as an isolated fact waiting to be confused with its neighbour.
Q: How can a candidate check whether a published section mapping chart is accurate?
Test the chart against anchors before trusting it anywhere else. Pick two correspondences whose positions are beyond argument, murder from Section 302 of the Penal Code to Section 103 of the Sanhita, and the electronic records certificate from Section 65B of the Evidence Act to Section 63 of the Bharatiya Sakshya Adhiniyam, and check whether the chart has them right. Then test its architecture: a chart that places an offence in the wrong neighbourhood of the code, a property offence among the offences against the body, has failed a structural check even where an individual number happens to be right. Finally, verify any row the answer will actually rely on against the bare text. Charts in wide circulation contain errors, and a candidate’s defence is a method, not a better chart.
Q: What should a candidate do when a question paper cites a section from the wrong code?
Answer the doctrine the question is reaching for, and show that the citation has been noticed. Identify the provision the examiner means from its content, state it correctly with its code and number, and proceed to answer on the merits, flagging the discrepancy in a line rather than building the answer on the error or silently adopting it. In an objective paper, choose the option that matches the doctrine described, since the options usually reveal the intended provision. Papers set during a transition can contain errors, and the skill being tested, whether the setter intended it or not, is recognition from content. The candidate who knows offences by their elements can do this; the candidate who knows only numbers is stranded by someone else’s mistake.
Q: How should a teacher sequence the old and new codes in a criminal law course?
Teach the doctrine once, in its structural home, and teach the transition as a separate, explicit unit. A course that runs the old code and then the new code as two complete passes doubles the load and teaches the correspondence by accident; a course that teaches elements, architecture and propositions first, with both numbers attached at the point of teaching, makes the correspondence the object of study. Reserve a distinct block for what genuinely changed, the new offences, the custody restructuring, the savings provisions, and give students the fault-finding exercise of checking a published mapping chart against the bare text. The sequence mirrors how the law itself is organised, and it leaves students able to work in the courts that still run on the old codes.
Q: How should a candidate use previous-year questions that were set under the old codes?
Use them as doctrine tests and re-solve them in both systems. A question framed on the Penal Code still tests elements, exceptions and propositions that carried over, so work it fully in the old code first, since that is the code its answer belongs to, and then perform the translation: the corresponding provision in the Sanhita, whether the law changed on the way, and how the answer would be cited in a paper set under the new code. The exercise converts an archive that looks obsolete into the best available training for transition papers, which are themselves built by setters translating their old questions. Discard from the archive only the questions whose substance was repealed outright, and note even those, because repeal is itself examinable.
Q: What should a one-line proposition note for a leading judgment contain?
It should contain the holding and nothing else: the court, the point decided, and the limit of what was decided, in one sentence that could be written into an answer without alteration. A note on Lalita Kumari that says the case was about first information reports is a label, not a proposition; the proposition is that registration is mandatory where a cognizable offence is disclosed, with preliminary inquiry confined to limited categories. Keep the bench and the year as retrieval hooks at the head of the note, keep observations and commentary out of the sentence itself, and test the file by covering the case names and stating the propositions from the facts. A proposition that cannot be produced cold is not yet learned.
Q: Should a candidate study the codes chapter by chapter or offence by offence?
Study the architecture chapter by chapter once, and the content offence by offence thereafter. The first pass through a code should follow its own order, because the order is information: general principles before offences, offences grouped by what they attack, procedure arranged as a sequence, evidence arranged as a logic. After that pass, revision by offence and doctrine is faster and matches how questions are set, since examiners ask about murder, bail or confessions and not about chapters. The error to avoid is beginning offence by offence with no map, which produces a candidate who knows provisions and cannot place them, and who therefore cannot do the one thing the transition rewards, which is to find an unfamiliar number from its neighbourhood.
Q: How much of an aspirant’s criminal law time should go to procedure and evidence rather than offences?
More than instinct suggests. Candidates gravitate to the substantive code because offences are vivid, but the procedural and evidentiary provisions carry a heavy share of the tested material: arrest and bail, the first information report, remand and default bail, the confession rules, presumptions and the certificate for electronic records are all standing topics, and the judgments that dominate the field, Lalita Kumari, Arnesh Kumar, Anvar and Arjun Panditrao among them, are procedure and evidence cases. A defensible division gives procedure and evidence together about half the criminal law allocation, weighted by the paper being attempted, since judiciary papers press procedure harder than the civil services do. The test of the division is coverage of the lattice, which is deliberately built across all three codes.
Q: What does not need to be relearned when moving from the old codes to the new ones?
The substance of most of the field. The elements of the standard offences, the general exceptions, abetment and common intention, the procedural sequence from information through investigation to trial and appeal, and the evidentiary logic of relevancy, proof, burden and presumption all carried over with consolidation rather than reinvention as the dominant technique. The case law carries over wherever the successor provision is substantially the same as the one it replaced, which covers most litigated ground. What needs learning is the numbering for the provisions that matter, a finite list of genuine changes, and the transitional rule. A candidate who inventories the inheritance honestly can spend the transition effort where the change actually is, instead of relearning theft.
Q: How is the transition between the old and new codes itself tested in examinations?
At three depths, and a candidate should prepare for each. The shallow test is a date: facts placing an offence on one side of 1 July 2024, with a distractor, a later registration or a later trial, pulling towards the wrong code. The middle test is a stage: a proceeding pending at commencement, where the candidate must assign substance and procedure separately and cite the savings provisions that do the assigning. The deep test is a reason: why the substantive line is constitutional, resting on the bar on ex post facto criminal law in Article 20(1), while the procedural line is a legislative choice expressed in the savings clauses. Papers reward the candidate who can move between the depths without confusing them.