A criminal case in India is shaped less by the offence it alleges than by the procedure that carries it, and on 1 July 2024 that procedure changed hands. The Bharatiya Nagarik Suraksha Sanhita, 2023, Act 46 of 2023, repealed the Code of Criminal Procedure, 1973 and took its place as the code that governs how information reaches the police, how investigation is conducted, how custody is authorised, how a trial is run and how a judgment is delivered. The Sanhita was advertised on two promises: speed, delivered through a dense network of statutory timelines, and modernisation, delivered through electronic process and forensic science. Whether those promises bind anyone is a different question, and it is the question this article tests provision by provision, because a procedural reform is only as real as the consequence that follows when it is ignored.

The distinction that organises everything that follows is between a duty and a direction. Much of what the Sanhita announces was already the law before it was enacted, stated by the Supreme Court in decisions that police forces and trial courts were expected to obey and frequently did not. Codifying a judicial direction into a statute changes its source; it does not by itself change the behaviour of the institution that was disregarding it. Other parts of the Sanhita are genuinely new law, and some of those change the position of an accused person in ways that matter from the first week of custody. Sorting the first category from the second, and then asking of every new period the statute prescribes whether anything happens when the period expires, is the only honest way to read this code.

Why the Code of Criminal Procedure Ended

The Code of Criminal Procedure, 1973, Act 2 of 1974, was itself a replacement statute, enacted on the foundation of the Law Commission’s Forty-first Report to displace the colonial code of 1898 and to give the republic a procedure of its own. The lineage matters to the 2023 replacement for one specific reason: the 1973 Code came to Parliament from the Law Commission, the standing body for law reform, while the 2023 codes came from an executive committee and a ministry, and the opposition’s procedural criticism in the passage debate fastened on exactly that difference, arguing that a generational recasting of criminal law deserved the Law Commission’s deliberative route and a wider public consultation than the bills received. The government’s answer, that the Committee for Reforms in Criminal Laws had consulted stakeholders extensively and that the standing committee stage supplied the parliamentary scrutiny, is also part of the record. The dispute is not decorative. It goes to how much testing the custody clause and the preliminary enquiry band received before enactment, and it explains why the courts, rather than the drafting process, are where those provisions are now being examined. It carried 484 sections and it organised the criminal process around institutions that the Sanhita has retained almost intact: the hierarchy of criminal courts, the distinction between cognizable and non-cognizable offences, the first information report as the trigger of investigation, the police report as its product, and the division of trials into sessions trials, warrant cases, summons cases and summary trials. A reader who knows the old code’s architecture will find the new code’s architecture familiar, because the replacement was drafted as a reorganisation of the inherited structure rather than as a new design.

What had failed, by the time the replacement was proposed, was not the architecture but the tempo. Criminal cases in the district courts ran for years at every stage: information resisted at the police station, investigation without a visible end, committal delayed, charges framed long after arrest, trials adjourned as a matter of routine, and judgments reserved and then delivered months after argument closed. The stock of pending criminal cases in the subordinate courts stood in the tens of millions, a figure that moves with every statistical release and is better understood as a standing condition of the system than as a number that can be quoted once and relied on. Undertrial prisoners, persons held in custody on accusations that had not been tried, made up a large and persistently criticised share of the prison population, and the Supreme Court returned to their position in proceeding after proceeding without the underlying pattern breaking.

The old code also carried a legitimacy problem that its defenders and its critics stated in different vocabularies. The government’s case, made in Parliament by the Home Minister, Amit Shah, was that the three criminal codes of 2023 closed the colonial chapter of Indian criminal law and re-centred the system on justice for the citizen rather than on the management of subjects. The critics of the old code who agreed with its repeal did not always agree with what replaced it, and the sharpest of them argued that a code can be colonial in technique as well as in origin if it enlarges police power while speaking the language of citizen protection. Both descriptions of the 1973 Code, as an outdated colonial inheritance and as a workable procedure weakened by the institutions operating it, are part of the record, and the replacement has to be judged against both.

The Making of the Sanhita

The immediate drafting history begins with the Committee for Reforms in Criminal Laws, constituted by the Ministry of Home Affairs and chaired by the Vice Chancellor of National Law University, Delhi, which submitted its report in February 2022 after consulting stakeholders across the system. The three replacement bills were introduced in the Lok Sabha on 11 August 2023 and referred to the Department-related Parliamentary Standing Committee on Home Affairs, which submitted its reports on the three bills, numbered 246, 247 and 248, on 10 November 2023, the report on this Sanhita being the 247th. The passage record from that point is shared with the substantive code and is treated in full in the companion account of the Bharatiya Nyaya Sanhita and its passage through Parliament: the first bills were withdrawn on 12 December 2023, revised bills were introduced the same day, the Lok Sabha passed them on 20 December 2023 and the Rajya Sabha on 21 December 2023, and the President assented on 25 December 2023.

The standing committee’s work on procedure deserves separate notice because it shows where scrutiny bit and where it did not. The committee examined the clause that became the custody provision of this Sanhita and recorded its apprehension that allowing police custody beyond the first fifteen days could be vulnerable to misuse, recommending a suitable amendment to give the clause greater clarity in interpretation. The revised bill retained the clause in its original form, so the committee’s most consequential procedural recommendation is also its clearest defeat, and the episode is a useful measure of how much of the final text the committee stage moved. On electronic records the committee recommended that the Sanhita provide for the secure handling and processing of electronic and digital records acquired in investigation, a recommendation that pointed in the same direction as the code’s own electronic-mode provisions. The Home Minister told Parliament that most of the changes made after the committee stage were grammatical in nature, and on the provisions that matter most to liberty, the enacted text bears that out.

One feature of the drafting process attracted a criticism independent of the committee stage: the Committee for Reforms in Criminal Laws had delivered its report to the Ministry in February 2022, but the report had not been published when the bills were introduced, so the reasoning from which individual clauses descended, and the consultations said to support them, could not be examined clause by clause against the drafts. A reform that arrives with its working unpublished invites the suspicion that the consultation was an exercise in collection rather than in persuasion, and the suspicion attached to the bills throughout their passage, fairly or not. The net growth of the code, forty-seven sections over its predecessor, also belongs to this account of the drafting: the increase understates the editing, because most of the new material, the forensic duty, the ninety-day information duty, the custody window and the adjournment limits, was written into inherited sections as clauses and provisos rather than added as new sections, so that the Sanhita’s novelty is concentrated inside familiar addresses, which is precisely why section-level comparison, of the kind this article has conducted, is the only reliable way to find it.

When did the Bharatiya Nagarik Suraksha Sanhita come into force?

The Bharatiya Nagarik Suraksha Sanhita received the President’s assent on 25 December 2023 and was brought into force on 1 July 2024. From that date it governs every investigation, inquiry and trial begun under it, while proceedings already pending on that date continue under the repealed Code of Criminal Procedure by force of the Sanhita’s own savings clause.

Commencement was not quite universal. The notification that brought the Sanhita into force excepted the entry in its First Schedule relating to Section 106(2) of the Bharatiya Nyaya Sanhita, the hit-and-run provision whose commencement was held back after protests by transport workers, a reminder that even a code enacted as a single event enters the world in pieces. The Sanhita runs to 531 sections arranged in thirty-nine chapters, against the 484 sections of the code it repealed, and the increase is itself informative: very little was cut, a great deal was carried across, and the additions cluster in the places this article examines, at the front of the process where information is received and in the middle where custody and trial are timed.

What the Bharatiya Nagarik Suraksha Sanhita Kept from the Old Code

The first analytical move in any replacement statute is to identify what survived, because continuity determines how much of the old learning remains usable. Under the Sanhita, the map of criminal courts is unchanged in substance: Courts of Session, Judicial Magistrates, Metropolitan Magistrates in the notified areas the old code recognised, Executive Magistrates with their preventive functions, and the High Courts above them. The classification of offences in the First Schedule, cognizable or not, bailable or not, triable by which court, follows the inherited pattern with the references renumbered to the Bharatiya Nyaya Sanhita. The stages of a case are the same stages. Information is recorded, investigation follows, the police report goes to a Magistrate, cognizance is taken, sessions cases are committed, charges are framed, evidence is recorded, the accused is examined, arguments are heard and judgment is pronounced. An advocate who learned the process under the old code does not have to learn a new process; the advocate has to learn new numbers and a set of specific changes laid over the familiar route.

The precedent carried across with the structure. Because Section 531 of the Sanhita repeals the old code while saving pending proceedings, and because most operative provisions reproduce their predecessors in substance, the case law built on the old code remains the starting point for reading the new one wherever the text is materially the same. The Supreme Court’s direction in Lalita Kumari v Government of Uttar Pradesh, a Constitution Bench decision, that registration of a first information report is mandatory when information discloses a cognizable offence, and its arrest-restraint directions in Arnesh Kumar v State of Bihar, which built a checklist and a notice-of-appearance discipline into the arrest provisions, are the two decisions this article returns to, because several of the Sanhita’s announced reforms are these directions restated. The full treatment of both decisions belongs to the complete guide to the Code of Criminal Procedure, which owns them for this series; here they matter as the baseline against which novelty is measured.

How many sections does the Bharatiya Nagarik Suraksha Sanhita have?

The Sanhita has 531 sections in thirty-nine chapters, replacing the 484 sections of the Code of Criminal Procedure, 1973. Most of the increase comes from additions rather than expansion of old provisions: new duties at the information and investigation stage, a chapter of timelines, and provisions on electronic process and forensic examination that the old code did not carry.

The functional map of the thirty-nine chapters confirms the continuity. The early chapters constitute the courts and their territorial divisions and house the prosecution service. The arrest chapters follow, then the processes to compel appearance and production, summons, warrant, proclamation and attachment. The preventive jurisdiction of Executive Magistrates occupies its inherited block, security proceedings, public nuisance, disputes over immovable property and preventive action by the police. The centre of the code is the sequence this article follows: information to the police and their powers to investigate; the jurisdiction of courts in inquiry and trial; the conditions requisite for initiating proceedings, which is where sanction and cognizance live; complaints to Magistrates; the commencement of proceedings before Magistrates; the charge; and then the four trial chapters, sessions trials, warrant cases, summons cases and summary trials, in descending order of gravity. Around the trial chapters sit the chapters on evidence in inquiries and trials, on judgment, on the submission of death sentences for confirmation, on execution, on appeals, references and revision, on the transfer of cases, and on bail and bonds, before the miscellaneous chapter and the closing chapter of repeal and savings. A reader laying the two codes side by side finds this map in both, in the same order, which is the structural fact beneath the argument about novelty: the replacement rearranged the furniture in a handful of rooms and rebuilt two of them, and left the house standing.

What was discarded is a shorter list than the rhetoric of replacement suggests. The Sanhita drops the old code’s metropolitan-area apparatus as a distinct administrative idea in the form the 1973 Code had built it, folds in provisions that had entered the old code piecemeal through decades of amendment, and abandons a small number of provisions that had fallen into disuse. The court structure itself illustrates the scale of the editing. Section 6 states the classes of criminal courts as the Court of Session, Judicial Magistrates of the first class, Judicial Magistrates of the second class and Executive Magistrates, and the metropolitan magistracy that the old code had maintained for the great cities does not appear as a separate class, its courts absorbed into the ordinary hierarchy. The sentencing powers distributed across that hierarchy are the inherited ones: a Chief Judicial Magistrate may pass any sentence authorised by law except a sentence of death, imprisonment for life or imprisonment for a term exceeding seven years; a Magistrate of the first class may pass imprisonment for a term not exceeding three years; a Magistrate of the second class, one year. Above them the Court of Session may pass any sentence authorised by law, with a death sentence subject to confirmation by the High Court, and the High Court any sentence at all. The distribution is why the classification of an offence, and the forum the First Schedule assigns it to, decides the practical ceiling on punishment in most prosecutions, and why the Sanhita’s retention of the structure carries with it the retention of the whole jurisprudence about which court may try what. The honest summary is that repeal here meant re-enactment with edits, which is precisely why the argument about the code cannot be settled by pointing at its newness. The argument has to be conducted provision by provision, and it divides into the provisions that restate, the provisions that time, and the provisions that genuinely move power from one actor to another.

Information and Investigation: the Codified Front End

The Sanhita’s most publicised changes sit at the point where a citizen first meets the system, and they are a study in the difference between invention and codification. Section 173, which corresponds to Section 154 of the old code, provides that information relating to a cognizable offence may be given to the officer in charge of a police station irrespective of the area where the offence was committed. That phrase writes the zero first information report into the statute. The practice was not new: the Justice Verma Committee, reporting in January 2013 after the Delhi gang rape, had recommended that a first information report be registered irrespective of the area of the offence, courts had directed for years that a station receiving information about an offence committed elsewhere must record it and forward it to the station with jurisdiction, and the Ministry of Home Affairs had advised the same course in 2015, particularly in offences against women. Police stations nonetheless continued to turn informants away on jurisdictional grounds, because a direction in a judgment and an instruction in a statute are enforced differently at a station-house desk. Section 173 gives the informant a text to point to, and that is a real change in the distribution of argumentative power, even though the idea itself is older than the code.

The same section recognises information given by electronic communication. An e-FIR given through electronic means is taken on record once the informant signs it within three days, and Section 173(2) entitles the informant or the victim to a copy of the recorded information free of cost, supplied forthwith. The signature condition is the detail that the publicity around electronic filing tends to omit: electronic information is a facility for starting the process from a distance, not a substitute for the informant’s attestation, and an unsigned electronic communication sits in a statutory waiting room until the signature arrives.

Against these two facilitative changes sits a restrictive one. Section 173(3) creates a statutory preliminary enquiry: where the information relates to a cognizable offence punishable with imprisonment of three years or more but less than seven years, the officer in charge may, with the prior permission of an officer not below the rank of Deputy Superintendent of Police, conduct a preliminary enquiry to ascertain whether a prima facie case exists, and the enquiry is to be completed within fourteen days. Under the old code, the Constitution Bench in Lalita Kumari had confined preliminary enquiry to a narrow set of categories, matrimonial and family disputes, commercial offences, medical negligence, corruption and cases of abnormal delay, and had held that in all other cases registration was compulsory on information disclosing a cognizable offence. Section 173(3) replaces that judge-made list with a punishment band, and the band covers a large share of the offences that citizens most commonly report. Defenders of the provision present it as a filter against mechanical registration of complaints that disclose no offence at all; critics answer that a filter operated by the police, at the threshold, on a fourteen-day clock, is the point at which inconvenient complaints quietly disappear. Both readings are available on the text, and the provision’s real character will be fixed by how courts police the prima facie enquiry it authorises.

What is the preliminary enquiry before an FIR under the BNSS?

Section 173(3) allows the officer in charge, with prior permission from an officer of Deputy Superintendent rank or above, to hold a preliminary enquiry where the reported offence is punishable with three years or more but less than seven years of imprisonment. The enquiry, which tests whether a prima facie case exists, must be completed within fourteen days.

The investigation provisions add two duties of a different kind. Section 176(3) addresses the quality of what investigation collects: on receipt of information about an offence punishable with seven years or more, the officer in charge must cause a forensic expert to visit the scene of the offence to collect forensic evidence, and must cause the process to be videographed on a mobile phone or another electronic device. The obligation is expressly phased. It operates from a date the State Government notifies within a period of five years, and where the forensic facility is not available for the offence in question, the State Government must notify the use of a facility in another State until its own is developed. The statute therefore concedes inside its own text that the mandate runs ahead of the infrastructure: India does not have, distributed across its districts, the forensic personnel and laboratories that universal scene-of-crime examination for every serious offence would require, and the provision is written to switch on State by State as capacity is built. The videography duty, by contrast, needs no laboratory and bites immediately as a matter of the officer’s own conduct, and the same is true of Section 105, which requires the process of search and seizure, including the preparation of the list of things seized, to be recorded through audio-video electronic means, preferably a mobile phone, with the recording forwarded without delay to the District Magistrate or the Sub-divisional Magistrate. The recording obligation lands on a search procedure whose human safeguards the code retains in full. Under Section 103, the successor of Section 100 of the old code, a search of a closed place is to be conducted in the presence of two or more independent and respectable inhabitants of the locality, the occupant is entitled to be present and to receive a copy of the seizure list signed by the witnesses, and a person found in the place may be searched on the conditions the section states. Those witnesses, the panch witnesses of Indian criminal practice, were the old system’s answer to the planting allegation: the recovery of a weapon or a packet was believed because disinterested neighbours had watched it happen. The answer decayed in practice into a formality of signatures, and Section 105’s recording is best understood as the technological successor to the panch system rather than as an ornament on it, an attempt to restore, by a device the officer already carries, the verifiability that the presence of neighbours was supposed to supply and often did not. The two safeguards now coexist, the witnesses and the recording, and the trial court of the Sanhita era will ordinarily have both a signed list and a forwarded video against which to test the recovery it is asked to act on.

The victim-facing duties complete the front end. In an investigation into a rape offence, Section 184 requires the medical examination of the victim to be conducted with consent by a registered medical practitioner and requires the practitioner to forward the report of the examination to the investigating officer within seven days, a period the old code did not impose on the corresponding provision. Section 193(3) requires the police to inform the informant or the victim of the progress of the investigation within ninety days, by any means including electronic communication. Neither duty carries a stated consequence, and both are best understood as standards against which conduct can be measured in court and in departmental proceedings rather than as self-executing rights. What forensic and electronic material is worth once collected, how it is proved, and what certificates it needs are questions for the evidence code, and they are treated in the companion account of the Bharatiya Sakshya Adhiniyam 2023.

The forensic duty deserves a further word on capacity, because the gap between the mandate and the means is the part of the reform most likely to be misreported in both directions. Forensic work in India runs through a thin institutional layer: a central Directorate of Forensic Science Services, a small number of central forensic science laboratories, State forensic science laboratories of very uneven depth, and district-level capacity that in many States amounts to a mobile unit and a photographer. Parliament created the National Forensic Sciences University by an Act of 2020 precisely to build the training pipeline the system lacked, which is an admission, in statutory form, of the shortage. Section 176(3) is written against that background. Its five-year notification window and its inter-State borrowing proviso are not loopholes discovered after the fact; they are the drafters’ own acknowledgment that a universal duty to send a forensic expert to every serious crime scene cannot be performed by the establishment that existed when the code commenced. The honest criticism is therefore not that the section is insincere but that it is a capacity-building programme wearing the clothes of a procedural right, and that an accused or a victim who invokes it in the first years of the code will usually be invoking a duty that the State has notified into force for some offences and some districts and not for others. The honest defence is that no previous code even attempted to make forensic collection a statutory incident of investigation, and that duties of this kind are how capacity arguments get won inside governments. Which description proves truer is an empirical question about notification and recruitment, and the article’s discipline is to leave it there rather than to decorate either side with figures.

Two boundary provisions complete the front end. For non-cognizable offences, Section 174, the successor of Section 155 of the old code, keeps the police out of investigation altogether unless a Magistrate orders it: the officer records the substance of the information, refers the informant to the Magistrate, and investigates only on the Magistrate’s direction, though the power to arrest follows the cognizable pattern once that direction is given. The distinction between the two classes of offence, which decides whether the State’s machinery starts on a citizen’s word or on a court’s, is among the oldest in Indian procedure and the Sanhita does not touch it. When the police will not move in a cognizable case, the Sanhita preserves the citizen’s ladder of escalation in the form the courts built under the old code: the application to the Superintendent of Police under Section 173(4), then the application to the Magistrate under Section 175(3), who may order registration and investigation or treat the application as a complaint, and behind both, the inherent powers of the High Court preserved by Section 528, the successor of Section 482 of the old code, to make such orders as are necessary to give effect to the code, prevent abuse of the process of any court or otherwise secure the ends of justice. Section 528 deserves its sentence in this account because it is the provision through which the High Courts filled the old code’s gaps for fifty years, quashing proceedings, ordering investigations and moulding relief the code had not imagined, and its re-enactment in identical terms is the largest single transfer of judicial authority in the whole replacement: Parliament re-enacted, knowingly, a blank authority whose contents are whatever the High Courts have said they are.

The witness-facing provisions of the investigation chapter were adjusted at the margins, and the adjustments run in the protective direction. Under Section 179, the successor of Section 160 of the old code, a police officer may require the attendance of a person acquainted with the facts of the case, but the proviso protects the vulnerable from the summons to the station: no male person under the age of fifteen years or above the age of sixty years, no woman, and no person who is mentally or physically disabled is to be required to attend at any place other than the place of residence, with attendance elsewhere possible at the person’s own option. The Sanhita lowered the protected male age from sixty-five to sixty and added disability to the protected categories, a small, precise widening of a provision that decides whether an elderly witness gives a statement at home or at a police station. Statements made to the police are governed by Section 180, the successor of Section 161, and their use remains confined by Section 181, the successor of Section 162: a statement recorded in investigation is not to be signed by the maker and may be used at trial only to contradict the witness in the manner the evidence code permits. That pair of provisions, more than any other in the chapter, preserves the investigation’s character as an inquiry rather than a proof, and the Sanhita leaves them in their inherited form. The statement of a victim of a sexual offence stands on a different footing, being recorded by a Magistrate under Section 183(6) as soon as the offence is brought to the notice of the police, with assistance for a victim who is temporarily or permanently disabled and with interpretation and videography safeguards, so that the most consequential account in the case enters the record through the court rather than the station.

Two investigation provisions complete the picture of the front end, one about who can start an investigation and one about what counts as its record. Section 175 governs the police power to investigate a cognizable case and the Magistrate’s power to order an investigation on a complaint, the route a citizen takes when the station will not move. The Sanhita tightens that route in a way that cuts in the opposite direction from the zero first information report: a Magistrate ordering investigation is to do so after considering the application, which the provision now expects to be supported by an affidavit, and after such inquiry as the Magistrate thinks necessary, a discipline borrowed from the Supreme Court’s directions against mechanical forwarding of complaints to the police. The same section empowers the Magistrate to monitor the investigation ordered, carrying into the text a supervisory practice the courts had developed. Section 183 governs the recording of confessions and statements before a Magistrate, the successor of Section 164 of the old code, and adds a facility with evidentiary weight: a confession or statement under the section may be recorded by audio-video electronic means in the presence of the advocate of the person accused. Confessions to the police remain inadmissible, and the Magistrate’s recording remains the channel through which a confession can reach the trial; the Sanhita’s addition means the voluntariness that the Magistrate must certify can now be tested against a recording made with defence counsel present, which alters the forensic character of the most contested document in a criminal trial. The police report itself, under Section 193, may be forwarded to the Magistrate through electronic communication, and the section’s further-investigation power, under which the police may continue investigating even after the report is filed and, where the trial has begun, are to complete that investigation within ninety days subject to the court’s power to extend the period, closes the investigation chapter with one more clock for the table’s argument.

The police report is not self-executing, and the Magistrate’s options on receiving it are the investigation chapter’s last piece of inherited machinery. The Magistrate may accept the report and take cognizance of the offence and issue process; may disagree with the conclusion the police have reached and drop the proceedings; or may order further investigation, the three courses the Supreme Court mapped in Bhagwant Singh v Commissioner of Police and which the Sanhita’s scheme preserves. Two protections travel with the options. Where the report recommends closure, the informant is entitled to notice and to be heard before the Magistrate accepts the closure, and the informant who disagrees may file a protest petition which the Magistrate may treat as a complaint and proceed with under the complaint chapter. And where the report discloses an offence, the Magistrate is not bound by the police’s grading of it: cognizance is of the offence the material discloses, not of the section the investigating officer has chosen. These doctrines are the working constitution of the transition from investigation to prosecution, they were built entirely by the courts on spare statutory text, and they survive the replacement for the reason given in the audit below: the Sanhita re-enacts the text they were built on.

The New Timelines Table

The Sanhita’s centrepiece claim is that it disciplines the process with time. The claim can be tested only by laying the periods side by side and asking of each the question the statute usually declines to answer: what happens if the period is missed? The table below is the article’s findable artifact, and its final column is the argument. A period with a specified consequence is a rule. A period without one is an instruction the institution gives itself, enforceable only to the extent that a court, in an individual case, chooses to treat the breach as a reason to intervene.

Stage	Provision	Period	Who is bound	Consequence if the period is missed
Signature on information given electronically	Section 173(1)	Three days	The informant	The information is taken on record on signature; no other consequence stated
Preliminary enquiry before registration	Section 173(3)	Fourteen days	The officer in charge, with Deputy Superintendent permission	None specified
Forensic expert visit for offences punishable with seven years or more	Section 176(3)	From the date notified by the State Government, within a five-year window	The officer in charge of the police station	None specified, and the duty itself is phased in
Medical examination report in a rape investigation	Section 184(6)	Seven days	The registered medical practitioner	None specified
Investigation of listed sexual offences	Section 193(2)	Two months from the recording of information	The officer in charge of the police station	None specified in the section
Progress of investigation to be told to the informant or victim	Section 193(3)	Ninety days	The investigating officer	None specified
Custody pending investigation before default bail	Section 187(3)	Sixty days, or ninety days for offences punishable with death, life imprisonment or ten years or more	The investigating agency	The accused is entitled to be released on bail on furnishing it; the one investigation-stage period with a written consequence
Supply of the police report and documents	Section 230	Fourteen days from the production or appearance of the accused	The Magistrate	None specified
Committal of a sessions case	Section 232	Ninety days from taking cognizance, extendable for recorded reasons to a maximum of one hundred and eighty days	The Magistrate	None specified
Application for discharge in a sessions case	Section 250	Sixty days from the date of commitment	The accused	The window governs the accused’s own application
Framing of charge	Sections 251 and 263	Sixty days from the first hearing on charge	The trial court	None specified
Judgment in a sessions trial	Section 258	Thirty days from the completion of arguments, extendable to forty-five days for reasons recorded in writing	The Court of Session	None specified beyond the condition on extension
Judgment, general outer limit	Section 392	Forty-five days from the termination of the trial	Every court of original jurisdiction	None specified
Decision on sanction to prosecute a public servant	Section 218	One hundred and twenty days from receipt of the request	The Government	Sanction is deemed to have been accorded; a written consequence
Trial of listed rape offences	Section 346	Two months from the filing of the chargesheet	The trial court	None specified; carried over from the old code

Two rows carry consequences, and they are the two rows where the consequence was the point of the provision rather than an ornament on it. Section 187(3) continues the default bail mechanism of the old code: if the investigation is not completed and the police report not filed within sixty days, or ninety days where the offence is punishable with death, imprisonment for life or a term of ten years or more, the accused who is prepared to furnish bail is entitled to be released, and decades of case law under the predecessor provision treated that entitlement as indefeasible once accrued. Section 218, examined below, attaches deemed sanction to governmental silence. Every other period in the table is silent about its own breach. The likeliest judicial treatment of those silent periods, on the pattern of the old code, is that they will be read as directory rather than mandatory: a charge framed on the sixty-first day will not fall, and a judgment delivered on the fiftieth day will not be void, because courts asked to void proceedings for breach of a consequence-free deadline must weigh the breach against the public interest in a trial on the merits, and they have historically declined to let the case die of a timetable. The periods will still do work. They give the accused, the victim and the appellate court a statutory yardstick for delay, they arm a High Court exercising superintendence, and they convert complaints about tempo from rhetoric into arithmetic. What they do not do is enforce themselves.

Default bail shows what enforcement looks like when a period is built to be enforced, because its doctrine is fully formed. The right is indefeasible once the period expires without a police report, but it must be availed of: the accused who applies for default bail and is prepared to furnish it before the report is filed has exercised the right, and a report filed thereafter, even on the same day, does not extinguish it, the principle the Supreme Court worked through in M. Ravindran v Directorate of Revenue Intelligence on the foundation of the earlier cases. The practical consequence is that the sixtieth and ninetieth days are dates on which both sides attend court with rival documents, the prosecution with a report, the accused with a bail application, and the order of filing decides the liberty of the person concerned. No other pair of dates in the code carries that charge, and the contrast states the article’s argument in its purest form: a clock matters exactly to the extent that someone can convert its striking into an order, and the test of every other period in the table is whether a court, a victim or an accused can find an equivalent conversion for it.

Can an accused get free copies of the police report and documents under the BNSS?

Yes. Section 230 requires the Magistrate to supply the accused, without delay and free of cost, with a copy of the police report, the first information report and the documents forwarded with the report, and it fixes a period of fourteen days from the production or appearance of the accused for that supply to be made.

The fourteen-day supply period matters beyond its own terms because of the document-heavy character of the defence it feeds. A chargesheet and its enclosures reach the accused as a bundle of statements, seizure memos, medical reports and electronic records, increasingly in scanned form. Practitioners who maintain the bundle as an indexed electronic file, and who track the Sanhita’s periods against the case diary in a spreadsheet, can open and check those workbooks in the browser through ReportMedic’s office file viewer for spreadsheet, document and presentation files on their own machine, which keeps the papers of a brief from having to be uploaded to a document service before they can be read. The tool is incidental; the discipline is not. A code that runs on periods rewards the advocate whose dates are exact.

The Clocks That Are Missing

The table’s silences are as instructive as its entries, because they show where the drafters’ appetite for periods stopped. There is no statutory period for a Magistrate to take cognizance after the police report arrives, although cognizance is the hinge on which investigation becomes prosecution and although files wait at that hinge in every district. There is no general outer limit for the duration of a trial: the two-month completion period in Section 346(1) is confined to the listed sexual offences, and a sessions trial for murder or a warrant case for cheating may lawfully run for as long as the adjournment discipline permits, which is to say that the code times the trial’s end, in the thirty and forty-five days allowed for judgment, but not its middle. There is no period for the disposal of a bail application, the single most time-sensitive application in criminal practice, whose urgency the courts have repeatedly described and the statute leaves unmeasured. There are no periods for appeal or revision in the code itself, the limitation for those remedies continuing to come from the Limitation Act as it did before. And there is no period within which the State must complete the one act on which the forensic mandate depends, the building of the facility, because the five-year window in Section 176(3) governs notification of the duty and not the existence of the means.

One period of the old code does carry a consequence, and the Sanhita retains it untouched, which completes the taxonomy. Section 514, the successor of Section 468 of the old code, bars a court from taking cognizance of an offence after the lapse of the limitation period the section prescribes: six months where the offence is punishable with fine only, one year where it is punishable with imprisonment for a term not exceeding one year, and three years where the term exceeds one year but does not exceed three years, with cognizance of the graver offences above that band not time-barred at all. The bar has teeth, since a cognizance taken after the period without condonation is liable to be quashed, and it is softened by the companion provision under which the court may take cognizance after the period where the delay has been explained or the interests of justice require it. Limitation is the standing proof that Parliament knows how to attach a consequence to a period when the period’s purpose is repose: the stale prosecution is barred because the point of the rule is that some accusations come too late to be tried fairly. The investigation and trial periods of the new code have a different purpose, the discipline of actors inside a live case, and their consequencelessness is the price of keeping the case alive. Reading the code whole, with Section 514 at one end and Section 218 at the other, the spectrum is complete: a period that kills the proceeding, a period that grants the permission, and between them the great middle of periods that merely measure.

This pattern is worth naming precisely, because it distinguishes two kinds of legislative caution. The drafters were willing to time the actors they could see as delinquent: the police, who investigate slowly; the government, which sits on sanctions; and, within limits, the trial court, which reserves judgments. They were not willing to time the system’s structural bottlenecks, cognizance, trial length and appellate disposal, where a deadline with a consequence would collide with the number of courts and judges the State has actually provided. A deadline for judgment is safe to enact because thirty days is within a single judge’s control; a deadline for trial would be a promise about thousands of courtrooms. The consequence-free character of the periods that were enacted is therefore not best read as drafting failure. It is the visible edge of what a procedure code can promise without a budget attached, and the reader who understands that will read the timelines table as a map of the possible rather than as a timetable of the expected.

Arrest and Adjournment: the Supporting Discipline

Two further pieces of the code’s machinery carry its argument about tempo and restraint, and both are codifications with one hard edge each. The first is arrest. Section 35, the successor of Section 41 of the old code, governs arrest without warrant. For cognizable offences punishable with up to seven years, arrest is not the default: the officer must have reason to believe the person committed the offence and must be satisfied that arrest is necessary for one of the purposes the section lists, preventing a further offence, ensuring proper investigation, preventing the disappearance or tampering of evidence, preventing the influencing of witnesses, or ensuring the person’s presence in court, and the reasons must be recorded in writing. Where arrest is not required, Section 35(3) requires the officer to issue a notice directing the person to appear, and a person who complies with the notice is not to be arrested unless reasons are recorded for concluding that arrest has become necessary. This is the Arnesh Kumar discipline, the checklist and the notice, written into the statute that the judgment had to construct out of the old code’s spare text. Its presence in the Sanhita illustrates the codification pattern at its most defensible: a direction that depended on a judgment’s authority now stands in the provision a station officer actually reads.

The hard edge is Section 35(7), which has no counterpart in the old code or in the judgment. No arrest may be made without the prior permission of an officer not below the rank of Deputy Superintendent of Police where the offence is punishable with imprisonment of less than three years and the person is infirm or is above sixty years of age. The provision moves a class of arrests, the elderly and the infirm accused of minor offences, from the station officer’s discretion to a senior officer’s recorded permission. It is a small rule with a precise trigger, and it belongs in any audit of the code because it shows the drafters could write a gated power when they chose to: the gate here is not a period but a permission, and it operates before the act rather than after it.

Process is the third piece, and the least visible. The chapters on processes to compel appearance carry the code’s other electronic change: a summons may be served by electronic communication, and service so effected is recognised as valid service, with the summons bearing the image of the court’s seal. Service of process was the quiet scandal of the old system, the stage at which cases aged without a hearing because a summons could not be shown to have reached its man, and electronic service, where the record of transmission is generated by the act itself, attacks the same delay from the opposite end to the timelines: not by commanding speed but by making proof of the first step cheap and durable. Combined with the electronic mode permitted by Section 530 for the proceedings themselves, the process chapters make the Sanhita’s electronic layer a matter of service and record no less than of hearings, which is the level at which technology changes procedure in practice.

The second piece is the adjournment discipline of Section 346, the successor of Section 309 of the old code. Section 346(1) directs that proceedings in every inquiry or trial be continued from day to day until all the witnesses in attendance have been examined, and its proviso requires the inquiry or trial in the listed sexual offences to be completed within two months of the filing of the chargesheet. Section 346(2) then closes the exits through which trials used to leak. An adjournment at the request of a party may be granted only where the circumstances are beyond the control of that party; even then, not more than two adjournments may be granted to a party, after hearing the objections of the other side and for reasons recorded in writing; the fact that the advocate of a party is engaged in another court is not a ground for adjournment; and where a witness is present but a party or the party’s advocate is absent or unready, the court may record the witness’s statement and pass orders dispensing with that party’s examination of the witness. Adjournment culture was the mechanism by which the old code’s absence of timelines was converted into decades, and Section 346 is the Sanhita’s attempt to legislate against the culture directly rather than around it. It has a consequence structure of a kind, not a sanction but a forfeiture: the party that is not ready loses the witness, and the advocate’s diary is declared irrelevant. Whether trial courts, carrying the board they carry, will hold parties to two adjournments is the empirical question on which the tempo of the new code will actually turn.

Police Custody Rebuilt: the One Change Everyone Argues About

Every replacement code has a provision that practitioners identify within minutes as the one that changes their working lives, and in this Sanhita it is Section 187. The background rule comes from the old code as the Supreme Court read it. Section 167 of the Code of Criminal Procedure allowed a Magistrate to authorise detention in police custody for a total of fifteen days, and in Central Bureau of Investigation v Anupam J Kulkarni the Court held that those fifteen days could be taken only within the first fifteen days of remand; after that window closed, further detention could only be judicial custody, in prison, beyond the interrogating agency’s reach. The rule had a clear operational consequence. An investigating officer who had not finished interrogation, or who discovered late in the investigation that custody was needed again (a recovered weapon to be confronted, a co-accused arrested weeks later), could not return to the Magistrate for police custody at all.

Section 187(2) of the Sanhita keeps the total at fifteen days and abolishes the window’s position. The fifteen days may now be authorised in the whole or in parts at any time during the initial forty days of the detention period where the outer limit is sixty days, or during the initial sixty days where the outer limit is ninety. The aggregate is unchanged; the distribution is transformed. The difference from preventive detention, custody ordered on executive satisfaction without any charge at all, is fundamental and is mapped in the comparison of preventive detention and ordinary arrest; Section 187 custody remains judicially authorised custody inside a pending investigation, which is what makes its timing a question of court control rather than executive choice.

The machinery around the provision is inherited and worth stating, because it is where the court’s control is actually exercised. A person arrested without warrant may not be detained beyond twenty-four hours without a Magistrate’s order, and the first authorisation of police custody requires the accused to be produced before the Magistrate in person; the Magistrate acts on the case diary, the record of the investigation’s steps that the officer must transmit, and is expected to satisfy himself that the grounds for custody exist rather than to register the application. The Sanhita preserves this structure without strengthening it. Nothing in Section 187 requires the Magistrate to record why custody is needed at a late stage of the investigation as opposed to its beginning, nothing distinguishes the first application from the fourth, and nothing obliges the court to ask what the earlier days of custody produced. The custody debate is therefore not really about the number fifteen, on which both codes agree, or even about the window, on which they differ. It is about whether a power to return for interrogation across forty or sixty days will be supervised as a series of fresh decisions, each needing its own justification, or administered as a balance to be drawn down. The text permits the first and does not compel it, and that silence, rather than the window itself, is what the standing committee was reaching for when it asked for clarity.

The case for the change, made by investigators and accepted by the government, is that the Anupam J Kulkarni window forced custody applications to be made defensively and at once, before the investigation knew what it needed, and that serious investigations, financial trails, encrypted devices, absconding co-accused, do not arrange their needs inside a fortnight. The case against, made by the defence bar and recorded in substance by the Parliamentary Standing Committee on Home Affairs when it sought a clarifying amendment, is that the threat of renewed police custody hanging over an accused for forty or sixty days changes the character of judicial custody, discourages bail applications in the early weeks because release may simply invite a custody application, and exposes the accused to interrogation at the point where the investigation has had time to build pressure. The textual point that sharpened the committee’s concern was an omission: where the old code said that detention beyond fifteen days must be otherwise than in police custody, the new provision does not repeat the phrase, and the committee recommended that the clause be clarified to state the circumstances in which custody beyond the first fifteen days is proper. The recommendation was not taken up.

The courts had, in fact, begun to move before Parliament did. In the litigation over the arrest of V Senthil Balaji by the Enforcement Directorate, the Supreme Court in 2023 doubted the Anupam J Kulkarni reading and referred its correctness to a larger bench, and a coordinate bench in Central Bureau of Investigation v Vikas Mishra permitted police custody beyond the first fifteen days under the old code itself. That sequence matters to the honest audit of the Sanhita: the custody restructuring is a genuine substantive change, the single clearest enlargement of investigative power in the code, but it is a change that ratified a direction the judiciary was already travelling rather than one imposed on a settled contrary doctrine. What the Sanhita has not settled is the standard a Magistrate should apply to a custody application made on the thirtieth or the fiftieth day: the text authorises, it does not guide, and the early High Court decisions on Section 187 have so far addressed its arithmetic, holding for instance that only the period of actual custody counts toward the computation, rather than the principles that should govern a late application. Until a larger bench or the Supreme Court in a Sanhita case fixes that standard, the provision remains what the brief for this article called it, contested rather than settled, and any account that presents it as resolved in either direction is ahead of the record.

Bail and the Undertrial: Continuity with Two Sharp Edges

The bail architecture is carried across with its numbers changed and its structure intact. Bail in bailable offences remains a right. Bail in non-bailable offences is governed by Section 480, the successor of Section 437 of the old code; the special powers of the High Court and the Court of Session sit in Section 483, the successor of Section 439; and anticipatory bail, the direction that a person apprehending arrest be released in the event of arrest, sits in Section 482, the successor of Section 438, retaining the bar on pre-arrest bail for the gravest sexual offences against children, with the references renumbered to Sections 65 and 70(2) of the Bharatiya Nyaya Sanhita. Default bail, the entitlement that arises when the investigation outruns its sixty or ninety days, continues under Section 187(3) as described above. A practitioner reading only the headlines of replacement would expect a new bail law; the text delivers the old bail law with one significant addition and one significant restriction, both in the undertrial provision.

The choice to leave bail law alone was made against a live invitation to do the opposite. In Satender Kumar Antil v Central Bureau of Investigation, the Supreme Court organised the offences governed by its arrest and bail directions into categories and suggested that the Union consider enacting a separate law on bail, on the model of the Bail Act in the United Kingdom, to remove the subject from the interstices of a procedure code. A replacement code was the obvious legislative opportunity to take up that suggestion, and the Sanhita declines it: bail remains where the old code kept it, distributed between the arrest provisions, the custody provisions and the bail chapter, under the tests the courts have built. Defenders of the choice can point to the continuity argument that follows; critics can answer that a Parliament willing to renumber five hundred sections was not deterred by bulk, and that the refusal to consolidate bail is a measure of how far the replacement’s ambition actually ran. Either way, the governing law for a bail applicant on the first day of the Sanhita was, in substance, the governing law on the last day of the old code.

The continuity is not a defect to be explained away; in bail law it is a form of protection. The tests that govern the grant of bail, the considerations of flight, tampering and repetition, and the Constitution Bench’s holding in Sushila Aggarwal v State (NCT of Delhi) that anticipatory bail is not to be confined to a fixed term as a matter of course, were built on the words of the old code, and because Sections 480, 482 and 483 reproduce the operative scheme of their predecessors, that jurisprudence crosses over without reargument. A replacement code that had rewritten the bail tests in new language would have bought novelty at the price of a decade of relitigation over questions that are settled. The Sanhita’s drafters, to their credit on this chapter, renumbered and left the doctrine alone. The two places where they did not leave it alone are the undertrial provision and, in the custody chapter, the default bail clock’s new neighbourhood, and both deserve the scrutiny the next paragraphs give them.

The provision has a lineage that explains its shape. The undertrial problem entered constitutional law with Hussainara Khatoon v State of Bihar, where the Supreme Court confronted prisoners who had spent longer awaiting trial than the maximum sentence for the offences charged, and read a right to a speedy trial into Article 21. Successive decisions repeated the directions without dismantling the problem, and Parliament’s statutory answer, Section 436A of the old code, arrived only with the amendment of 2005 and capped detention at one-half of the maximum sentence, subject to exceptions and to the court’s power to order continued detention for reasons recorded. The undertrial majority in the prisons, the condition the Court had described across those decades, is the standing background against which Section 479 must be read: it is a provision enacted for a population whose detention is administrative routine rather than judicial decision, and its changes, the one-third rule for first-time offenders and the multiple-cases bar, operate on that population directly. The addition is in Section 479, which governs the maximum period for which an undertrial prisoner can be detained. The old code’s corresponding provision, Section 436A, required release on bail once an undertrial had undergone detention up to one-half of the maximum imprisonment specified for the offence, offences punishable with death excepted. Section 479 widens the exclusion to offences punishable with death or life imprisonment, and then adds the genuinely new rule: a first-time offender, a person never previously convicted of any offence, is to be released on bond on reaching one-third of the maximum period. Section 479(3) reinforces the mechanism by requiring the superintendent of the jail to apply to the court for the prisoner’s release when the half or the third is completed, converting a right that depended on the prisoner finding a lawyer into a duty owed by the custodian. For an accused in that position, access to counsel and to the legal services machinery is often the whole case, and the framework that supplies it is the subject of the account of the Legal Services Authorities Act 1987.

The restriction is Section 479(2). Where an investigation, inquiry or trial in more than one offence, or in multiple cases, is pending against a person, the section directs that the person shall not be released on bail under the undertrial provision at all. The carve-out has attracted the heaviest criticism the bail chapter has received, because the undertrial population is precisely the population against whom multiple cases are commonly pending, and a benefit drafted for first-time offenders and long-held undertrials can be switched off by the pendency of a second case of whatever weight. Defenders of the clause answer that the old provision was read to permit release in one case while others stood unresolved, producing releases the system then had to reverse; critics answer that the cure abolishes the benefit for the class that needs it. The provision is stated here in its terms because its terms are the debate. The general position described in this article is legal information about how the code is written, not advice on any individual case, and custody and limitation questions in a live matter turn on facts that only the case record shows.

Trial in Absentia, Deemed Sanction and the Remaining Additions

Section 356 introduces a procedure the old code did not have in this form: the trial of a proclaimed offender in his absence. A person declared a proclaimed offender under Section 84 who has absconded to evade trial, with no immediate prospect of arrest, is treated by the section as having waived the right to be present, and the court may proceed with the trial and pronounce judgment as if he were present. The gateway to the procedure is the proclamation itself. Under Section 84, where a court has reason to believe that a person against whom a warrant has issued is absconding or concealing himself so that the warrant cannot be executed, it may publish a written proclamation requiring him to appear at a specified place and time not less than thirty days from publication; and for the graver class of offences, those punishable with ten years or more, imprisonment for life or death, the court may, after making the inquiries it thinks fit, pronounce the person a proclaimed offender, the status Section 356 then acts on. The property machinery follows the status: the court may attach the proclaimed person’s property, and Section 86 adds a reach the old code lacked, providing for the identification and attachment of a proclaimed offender’s property held outside India through a request to the competent authority of the country concerned, an answer to the familiar pattern in which the proceeds of an economic offence are parked in a jurisdiction the Indian court’s warrant cannot touch. The procedure is fenced with conditions. The trial may not commence until ninety days have lapsed from the framing of the charge; two consecutive warrants of arrest must have issued at an interval of at least thirty days; notice must be published in a newspaper circulating at the place of the offender’s last known address, with thirty days to appear; a relative or friend must be informed; notice must be affixed at the offender’s house and displayed at the police station; and an offender who is not represented must be provided an advocate at the expense of the State. The case for the provision is the case against tactical absence: trials, particularly in economic offences, had been stalled for years by accused persons who placed themselves beyond arrest while their assets and their proceedings waited. The case against is that a criminal trial is built on confrontation, instruction of counsel and the accused’s own examination, and a judgment pronounced against an absent person, however carefully the notices were affixed, tests how much of that structure is dispensable. The provision is confined to proclaimed offenders, which is its limiting principle, and its constitutional testing belongs to the litigation the code is still generating.

The right of appeal against such a conviction is itself re-engineered, and the re-engineering may be tested before the trial procedure is. The proclaimed offender may appeal only if he presents himself before the appellate court, and no appeal lies after the expiry of three years from the date of the judgment, so the conviction hardens into finality unless the fugitive surrenders his freedom of movement to challenge it. The design is intelligible on the government’s terms: a man who has made the court’s process impossible cannot claim the process as a right, and the three-year window keeps the challenge open to the absconder who returns promptly while closing it to the one who waits out the witnesses. It is equally intelligible on the critics’ terms: the condition attaches a penalty, the loss of appellate review, to the very absence the trial was constructed to cure, and a conviction recorded without the accused ever being confronted with the evidence may stand unexamined although he returns in the fourth year with an answer to it. Between those two statements of the scheme there is little middle ground, and Section 356 will be read in the constitutional courts in the terms of one or the other.

Section 218, the successor of Section 197 of the old code, protects Judges, Magistrates and public servants from prosecution for acts done in the discharge of official duty without the previous sanction of the competent government, and it adds the second written consequence in the code. The government must decide a request for sanction within one hundred and twenty days of receiving it, and if it fails to decide, the sanction is deemed to have been accorded. The device answers a documented pathology: sanction requests under the old code waited for years, and the Supreme Court, in decisions including Subramanian Swamy v Manmohan Singh, had urged time-bound disposal without being able to write a deadline into a statute that did not contain one, a position the Court restated in Suneeti Toteja v State of Uttar Pradesh when it confirmed that no deemed sanction could be read into the old code. Parliament has now written the deadline in, and written the consequence in with it. Of all the timelines in the Sanhita, this is the one whose breach changes the legal position automatically, and the contrast with the silent periods in the timelines table is the cleanest evidence that the silence elsewhere was a choice rather than an oversight.

The Court’s own earlier attempt at a timeline throws the choice into relief. In Vineet Narain v Union of India, the Court laid down that a sanction decision should be taken within three months, with an additional month where consultation with the Attorney General was required. The guideline bound by expectation rather than by consequence, and the pathology it addressed survived it, which is why the Court in Subramanian Swamy addressed the recommendation to Parliament rather than to the sanctioning authorities. The comparison between the guideline and the statute is exact: the Court allowed three months with a possible fourth and made the remedy for breach a complaint; the code allows one hundred and twenty days and makes the remedy the grant itself. The section also preserves the scheme’s existing exceptions, including the disapplication of the sanction requirement for the offences the provision specifies, so the shield never extended to the classes of accusation Parliament had already decided should proceed without it. A government that wishes to refuse sanction under the new code must therefore do the one thing the old system was designed to avoid: record the refusal, within a period, and defend it.

The remaining additions are smaller and mostly point one way, toward a process conducted on screens and recordings. Section 530 provides that all trials, inquiries and proceedings under the Sanhita, including the issuance, service and execution of summons and warrants, the examination of the complainant and witnesses, the recording of evidence and appellate proceedings, may be held in electronic mode by the use of electronic communication or audio-video electronic means. Section 43(3) writes handcuffing into the code for the first time, permitting an officer, keeping in view the nature and gravity of the offence, to use handcuffs when arresting or producing in court a person who is a habitual or repeat offender, who has escaped from custody, or who has committed offences of organised crime, a terrorist act, drug-related crime, illegal possession of arms and ammunition, murder, rape, acid attack, counterfeiting of coins and currency notes, human trafficking, sexual offences against children or offences against the State; the provision sits in visible tension with the Supreme Court’s older handcuffing jurisprudence, which treated routine handcuffing as an affront to dignity requiring recorded justification, and the reconciliation of the text with that case law is another of the code’s open questions. Section 283 widens summary trial, empowering Magistrates, after hearing the accused and recording reasons, to try summarily offences not punishable with death, life imprisonment or a term exceeding three years, where the old code stopped at two years, and raising the property-value ceiling for the listed theft offences from two hundred rupees to twenty thousand rupees. And Section 360, the withdrawal-from-prosecution provision, requires the court to hear the victim before a prosecution is withdrawn, a small provision that redistributes a decision which the old code had left to the prosecutor and the court alone.

The Citizen Who Is Not Accused

Procedure codes are usually read from the accused’s side of the file, but the Sanhita also redistributes small rights among the people who appear in a case without being on trial in it, and an audit that ignores them underrates the code. The witness receives the benefit of the recording and electronic provisions: examination may be conducted by audio-video means, statements in the specified classes of cases may be recorded at the witness’s residence or place of choice, and the search and seizure that a householder endures is now a recorded event rather than a remembered one. The victim receives more. Beyond the free copy of the information and the ninety-day progress duty already described, the victim of an offence holds the right to be heard before a prosecution is withdrawn under Section 360, a right to have the medical examination in a rape investigation conducted and reported within the code’s own short periods, and the benefit of the compensation architecture the old code built and the Sanhita retains, under which a court may recommend compensation and the State’s victim compensation scheme, operating through the legal services authorities, supplies the fund where the offender cannot. None of these provisions was at the centre of the parliamentary argument, and each of them will be invoked in ordinary cases long after the contested five have been settled by the courts, which is a reason for an assessment of the code to weigh the unglamorous chapters alongside the famous ones.

The Quiet Bulk of the Code

A replacement statute is judged in the headlines by its five contested provisions, but it is used, day to day, in the bulk that nobody headlines, and the bulk of this code is re-enactment with administrative tidying. The prosecution machinery continues through Section 20, which retains the Directorate of Prosecution that the old code had inserted in 2005, with a Director of Prosecution and Deputy Directors supervising the conduct of prosecutions in the High Courts and the districts, a reform whose promise has always depended on whether the directorate is staffed as a cadre or as a posting. Cognizance continues through Section 210, the successor of Section 190 of the old code, and the complaint route continues through Section 223, under which the Magistrate examines the complainant and the witnesses present on oath before deciding whether process should issue, the filter through which private complaints have always passed. The maintenance jurisdiction that sends more citizens into Magistrates’ courts than any other single provision of the code continues in Section 144, the successor of Section 125, with its entitlement and its machinery substantively intact, and with it the Supreme Court’s guidelines in Rajnesh v Neha, which organised the overlapping maintenance claims arising under the personal laws and the code into a single framework of disclosure and assessment; the guidelines were framed on the old provision and pass to its successor without translation because the words that carry them were re-enacted. The urgent-order power of the Executive Magistracy continues in Section 163, the successor of Section 144 of the old code, under which a Magistrate may direct a person to abstain from an act likely to cause a breach of the peace or danger to human life, for a period of two months extendable by the State Government to six; the provision’s familiarity should not disguise its reach, since it is the instrument with which processions, festivals, construction and protest are regulated across the country, and the Sanhita’s decision to re-enact it without the procedural refinements its critics have proposed for decades is another instance of the replacement preserving the powers it might have been expected to discipline. The preventive chapters, under which Executive Magistrates take security for keeping the peace and for good behaviour, continue in their accustomed place, and the compounding of offences continues in Section 359, the successor of Section 320, with its two tables of offences compoundable by the victim, with or without the court’s permission.

The sentence end of the code carries a clock of a different moral weight. Where a death sentence is confirmed, Section 472 regulates the mercy petition to the President or the Governor with a sequence of periods for the first time in the code’s history: the convict, a legal heir or a relative may petition within thirty days of being informed by the jail superintendent of the dismissal of the appeal or special leave petition by the Supreme Court, the petition goes first to the Governor and then to the President within the periods the section prescribes, the Central Government is to recommend on the petition within sixty days of receiving the State Government’s comments and the jail records, and the President’s order is beyond challenge in any court. Under the old code the mercy process had no statutory timetable at all and moved at the pace of the secretariats, with delay itself becoming the ground on which death sentences were commuted. Section 472 is therefore the timelines argument carried to its extreme case: a deadline for clemency, with the consequence of lapse supplied by finality rather than by sanction. It was among the provisions most sharply criticised in commentary on the new codes, precisely because the consequence falls on the person with the most to lose from the State’s own pace, and it completes the article’s audit of clocks by showing that the drafters could attach a real consequence to a period when the period disciplined the citizen rather than the State.

Two items in the quiet bulk carry small clocks of their own, which is why they belong in this article’s audit. Plea bargaining, the chapter under which an accused in an eligible case applies for a negotiated disposition, sits in Sections 289 to 300, and the Sanhita adds a period the old code did not impose: the application is to be filed within thirty days of the framing of charge. The change is minor and its direction is characteristic, converting an open-ended facility into a window, on the theory that a bargain struck years into a trial is a failure of the trial rather than an alternative to it. Probation continues in Section 401, the successor of Section 360, under which courts may release a convicted person on probation of good conduct or after admonition instead of sentencing at once. Neither provision was debated in the replacement controversy, and both will affect more cases in an ordinary month than trial in absentia will in a decade, which is the standing reason for reading a procedure code whole rather than by its press coverage.

The schedules deserve their own sentence in a whole-code reading. The First Schedule classifies the offences of the Bharatiya Nyaya Sanhita for cognizance, bail and forum, and its second part grades offences under every other law by the same measures, so that an offence under a special statute enters the code’s machinery, its arrest powers, its bail classification and its trial forum, by the grade the schedule assigns to its punishment. The second Schedule carries the forms, the summons, the warrants, the bonds and the charges in their statutory words. Practitioners live in the First Schedule more than in any chapter of the code, and its re-enactment in step with the new offence numbers of the Sanhita is the mechanism by which the three replacement codes interlock: the Sanhita’s machinery is indifferent to the source of an offence, provided the schedule has graded it, which is why the simultaneous commencement of the three codes mattered and why a change in the numbering of the offence code had to arrive with its matching procedural schedule or not at all.

One further pair of provisions belongs to the citizen’s account of the code, because they state the ground rule of the adversary process. Section 340, the successor of Section 303 of the old code, provides that a person accused of an offence before a criminal court, or against whom proceedings are instituted under the code, may of right be defended by an advocate of his choice. Section 341, the successor of Section 304, supplies the counterpart where choice is empty for want of means: where the accused appears not to have sufficient means to engage an advocate in a trial before a Court of Session, the court shall assign an advocate for the defence at the expense of the State, and the High Court’s rules and the State Government’s provision extend the scheme as the section permits. The assignment machinery connects with the legal services authorities, which supply the panels and the payment, and it is the point at which the code’s promise of a fair contest is underwritten by money. The Sanhita re-enacts both provisions in substance, and their presence is a reminder that the most consequential sentences in a procedure code are sometimes the oldest ones in it.

After Judgment: Appeals, Revision and the Curative Chapters

The chapters that follow judgment explain why the replacement could be conservative without being inconsequential. Appeals under the Sanhita follow the inherited map: from a Magistrate’s conviction to the Court of Session, from a sessions conviction to the High Court, with the High Court’s confirmation required before a death sentence is executed and with the accused entitled to a copy of the judgment without delay. Section 413, continuing Section 372 of the old code with its proviso, preserves the victim’s right to appeal against an acquittal, against a conviction for a lesser offence, or against the imposition of inadequate compensation, a right that exists alongside the State’s appeal against acquittal and is not controlled by it. Revision, the High Court’s and the Sessions Judge’s power to call for the record of a lower court and correct a material irregularity, continues with its inherited limits, including the bar on converting an acquittal into a conviction in revision and the bar on revisional interference with interlocutory orders, the two limits that keep revision a supervisory remedy rather than a second appeal. The transfer jurisdiction, under which the Supreme Court, the High Court and the Sessions Judge may move a case between courts in the interests of justice, continues in its inherited breadth. And the closing chapters carry the curative provisions without which no code of this size could function: proceedings are not vitiated by irregularities of the kinds listed unless a failure of justice has been occasioned, a defect in the charge is material only if the accused was in fact misled, and findings are not reversed on appeal for errors that did not affect the result. These provisions decide more real cases than the headline chapters, because irregularity is the normal condition of litigation, and the Sanhita’s decision to carry them forward untouched is the deepest of its conservative choices: the new code was not prepared to destabilise the case law on what error matters.

Savings, Transition and the Two-Code Courtroom

Section 531 repeals the Code of Criminal Procedure, 1973 in terms and then saves its operation for everything already alive. Any appeal, application, trial, inquiry or investigation pending immediately before 1 July 2024 is to be disposed of, continued, held or made in accordance with the old code as it stood before commencement, as if the Sanhita had not come into force. Notifications, proclamations, powers, forms, jurisdictions, sentences and orders made under the old code are deemed to have been made under the corresponding provisions of the new one, and sanctions and consents accorded under the old code on which no proceeding had begun support proceedings under the new. Where the period for an application under the old code had already expired before commencement, the Sanhita’s longer periods do not revive it.

The practical consequence is a long dual regime. The Bombay High Court, in one of the earliest transition decisions, held that an investigation begun on a first information report registered before 1 July 2024 was a pending investigation saved by Section 531 and had to conclude under the old code, while applications filed on or after that date, including bail applications, are governed by the Sanhita because the old code stands repealed for everything not saved. Two procedural worlds therefore run side by side in the same courtroom, sorted not by the date of the offence or the date of the hearing but by the date the proceeding began: a 2023 investigation proceeds under the Code of Criminal Procedure while a 2024 investigation in the next file proceeds under the Sanhita, with different custody rules, different copy-supply periods and different section numbers for the same acts. The arrangement is the orthodox way to replace a procedure code, since procedure in pending matters is a vested course rather than a fresh right, but its duration is set by the pendency problem itself. As long as pre-commencement investigations and trials survive, and some will survive for years, the repealed code remains a living text that every practitioner must keep open beside its replacement.

The working courts have had to settle a set of practical questions the savings clause does not answer in terms, and the pattern of their answers is itself instructive. Citation practice settled first: a pleading or a judgment dealing with a saved proceeding cites the old code, and one dealing with a new proceeding cites the Sanhita, and where a judgment construes an old provision that the Sanhita reproduces, the ruling is cited for the new provision through the correspondence, because the words construed are the same words. Interlocutory questions have been harder. An application that is itself a fresh proceeding, a bail application filed after commencement in a case under investigation when the code changed, a revision against an order passed after commencement, invites the new code by the Bombay High Court’s reasoning, because the application is a proceeding commenced on the day it is filed, whatever the age of the case that carries it. A step inside a saved proceeding, an application for further investigation in a saved trial, a custody question inside a saved investigation, invites the old. The line that emerges is between proceedings and steps within proceedings, and it will be litigated at the margins for as long as saved cases last, which in practice means that the courts will be glossing Section 531 through the whole first generation of the Sanhita’s life. Practitioners have a further, purely mechanical burden during the same period: every precedent in the reports is cited to the old numbering, so that a lawyer arguing the corresponding provision of the Sanhita must carry the correspondence in her head or in her file, and the first commentaries on the new code are, in large part, exercises in mapping the old learning onto the new numbers.

The boundary has one deliberate softening, and it is on the side of liberty. In its continuing proceedings on prison conditions, the Supreme Court treated Section 479, the undertrial provision, as a beneficial provision whose benefit should reach undertrial prisoners whose cases began under the old code, rather than confining the one-half and one-third rules to prisoners arrested after commencement. The treatment is consistent with the way beneficial procedure is usually read on a change of code, and it stands in instructive contrast with the custody provisions, which no court has suggested applying backwards to investigations the old code governs. Transition law of this kind is unglamorous and decisive: it determines, for the large population of cases that straddle 1 July 2024, which of the two settlements of the custody and bail questions a given prisoner actually lives under, and it will keep doing so until the saved proceedings are exhausted.

The continuity that survived repeal runs deeper than the savings clause. Because most of the Sanhita re-enacts its predecessor in substance, the interpretive case law of the old code travels with the words: a provision reproduced without material change carries its settled construction with it, and courts reading the Sanhita have treated the predecessor jurisprudence as the natural starting point. Where the text departs, the departure is the argument. That principle explains why the genuinely contested litigation under the new code concentrates exactly where this article has concentrated, on the provisions whose words moved: custody in parts, the preliminary enquiry band, deemed sanction, trial in absentia, and the undertrial carve-out.

The Trial, Stage by Stage, Under the New Numbers

Because the Sanhita preserves the trial architecture, its stages can be stated once for all three of the trial forms it continues, and the statement is worth making in full, since the timelines table times only some of the joints. A trial opens with the charge: the court reads and explains the charge to the accused, who pleads guilty or claims to be tried. In sessions cases the discharge application under Section 250 precedes the charge, and in warrant cases the discharge application under Section 262 does the same work before a Magistrate; both are the accused’s argument that the material discloses no case to answer, and both now run on the sixty-day clocks recorded in the table. The prosecution evidence follows, witnesses examined in chief, cross-examined and re-examined, with the court empowered to summon and examine any person whose evidence appears essential to the just decision of the case. Then comes the examination of the accused under Section 351, the successor of Section 313 of the old code, in which the court questions the accused personally on the circumstances appearing in evidence against him, without oath and without liability to punishment for refusing to answer, so that he may explain what the prosecution’s witnesses have said. The defence follows, with its own witnesses and, in sessions trials, the accused’s entitlement to argue; the prosecution replies; and the court delivers judgment of acquittal or conviction, whereupon, on a conviction, a separate hearing on sentence precedes the order. In summons cases the sequence is compressed: the substance of the accusation is stated to the accused, there is no formal charge, and the court proceeds directly to hearing both sides. In petty cases the code continues the old facility under which an accused served with a special summons may plead guilty without appearing, transmitting the plea and the specified fine. Two features of the trial scheme deserve separate note because they carry the prosecution’s side of the architecture. In every trial before a Court of Session the prosecution is conducted by a Public Prosecutor, and the victim’s engagement of private counsel operates under the court’s control and subordinate to the prosecutor, a distribution of roles the Sanhita retains and which decides who actually speaks for the State in the trials that matter most. And the code retains the approver mechanism, under which the court may tender a pardon to a person supposed to have been directly or indirectly concerned in an offence, on condition of a full and true disclosure of the whole of the circumstances within his knowledge, converting an accomplice into a prosecution witness whose evidence is then tested by the special caution the evidence law applies to it. None of this is new law, and all of it is the law that governs every trial the new code will run, which is why the accurate mental model of the Sanhita is a familiar sequence of stages with a dated joint inserted at the points marked in the table and an argument, set out above, about what a date without a consequence can and cannot do.

The inquest provisions belong to the same anatomy, at its beginning rather than its end. Under Section 194, the successor of Section 174 of the old code, a police officer receiving information of a suicide or of a death in the circumstances the section lists inquires into the apparent cause of death and reports to the District Magistrate, and under Section 196, the successor of Section 176, a Magistrate inquires into the cause of death in the graver class of cases, including death or disappearance in police custody and the death of a woman within seven years of marriage in the circumstances the section describes, with the body forwarded for examination in aid of the inquiry. The Sanhita carries these provisions forward in substance, and their presence in an account of the replacement is a reminder that a procedure code governs the State’s first response to a death as completely as it governs a trial, and that on this part of the file the replacement changed almost nothing at all.

One Case Through the New Code

The difference the Sanhita makes is easiest to see by walking a single ordinary case along its route and noting where the ground has moved. Take an information of robbery, an offence punishable with ten years, given at a police station across the district from the place of the offence. Under the old code as practised, the first contest was jurisdictional: whether the station would record the information at all. Under Section 173 the information is recorded where it is given, as a zero first information report, and transferred; if it arrives electronically, the informant signs within three days and takes a free copy forthwith. Because the offence is punishable with seven years or more, Section 176(3) sends a forensic expert to the scene, where the State has notified the duty into force, and the collection is videographed; the search of the suspect’s premises, when it comes, is recorded under Section 105 and the recording goes to the District Magistrate. The investigation of the offence is not one of the two-month cases, but the victim is entitled under Section 193(3) to be told the progress of the investigation within ninety days, a date that can be diarised and demanded.

If the suspect is arrested, the custody negotiation that follows is where the new code bites hardest. The Magistrate may give the police up to fifteen days of custody, and the application need not exhaust them at once: five days now, and, if a co-accused is arrested on the fortieth day and a confrontation is needed, a further application within the sixty-day window that a ninety-day investigation period allows, since robbery carries ten years. Under the old code that second application was incompetent after the fifteenth day; under the Sanhita it is routine, and the defence argument has to be about justification rather than jurisdiction. If the police report is not filed by the ninetieth day, the accused is entitled to default bail under Section 187(3), and that entitlement, alone among the clocks in the case, enforces itself at the instance of the accused. When the report is filed, the Magistrate must supply the accused with the report and the documents free of cost within fourteen days under Section 230; because robbery is triable by a Court of Session, the case must be committed within ninety days of cognizance under Section 232, the accused may seek discharge within sixty days of commitment under Section 250, and the charge must be framed within sixty days of its first hearing under Section 251. At the trial, Section 346 holds the parties to day-to-day proceedings and two adjournments each; at its end, Section 258 gives the Judge thirty days from the completion of arguments, extendable to forty-five for recorded reasons, and Section 392 sets the forty-five-day outer limit in any event. Read end to end, the case has acquired a date at every joint where the old code had a practice. Whether the dates hold is the question the audit below takes up; that they exist to be invoked, by the victim at day ninety of silence, by the accused at day fifteen of document delay, by either party at the third requested adjournment, is the code’s working difference from its predecessor.

The same case, run for the accused rather than against him, shows the other set of dates. If he is a first-time offender and the trial outruns its clocks, Section 479 entitles him to release on bond when he has undergone one-third of the maximum period of imprisonment specified for the offence, provided no other case is pending against him, and the jail superintendent is bound by Section 479(3) to move the application when the period completes, so the entitlement does not depend on his having counsel alert to the arithmetic. If he is convicted and appeals, the appeal lies to the High Court from the sessions conviction, and the appellate chapters he enters are the inherited ones, with their powers to suspend the sentence and grant bail pending appeal. If he is acquitted, the victim and the State each hold their own right of appeal under the provisions already described, so the case is not necessarily over at the acquittal either. Walked in both directions, the ordinary case is the right unit of assessment for this code: not the constitutional challenge that will define one section, but the dated, documented sequence that five hundred other unnamed cases will follow every week, and on which the Sanhita’s real influence on Indian criminal justice will be exerted.

Codification or Innovation: the Honest Audit

The reform account of the Sanhita, stated at its strongest, is the government’s. The Home Minister presented the three codes of 2023 as the closing of a colonial settlement and the installation of a citizen-centred process, and on the face of this code the account can point to real entries: a statutory zero first information report that a station officer can no longer refuse on jurisdictional grounds; information receivable electronically; a forensic duty for serious offences with videography; a ninety-day duty to tell the victim how the investigation is progressing; a fourteen-day duty to hand the accused the papers of the case; a first-time undertrial’s release at one-third; a sanction regime in which official silence no longer buries a prosecution; and a full chapter of periods fixing how long each stage of a case is supposed to take. A code that writes these things down gives citizens, victims and accused persons texts to invoke, and the history of the old code shows that invoked texts are how Indian criminal procedure has usually been moved.

The dismissive account, also stated at its strongest, is that the Sanhita is the old code renumbered and retitled, its announced novelties largely restatements of what courts had already ordered, its timelines unenforced and unenforceable, and its one unmistakable transfer of power running in favour of the police through the custody clause. On this account the zero first information report codifies a direction from the judgment era, the forensic mandate switches on only when a State notifies capacity it does not have, the periods in the timelines table carry no consequence because the drafters did not intend them to disturb a single conviction, and the preliminary enquiry hands back at the police-station door part of what Lalita Kumari had taken away. The account is not refuted by pointing to the code’s length or its new name, and any honest assessment has to concede its strongest exhibits.

The verdict this article defends is that both accounts are true of different provisions, and that the code has to be read in the three layers the evidence supports. The first layer is restatement: zero information reports, recorded searches, arrest discipline and much of the victim-facing apparatus codify judicial directions, and their value depends on whether a statutory text is harder for a station officer to ignore than a judgment was, which is plausible but unproven. The second layer is the clock: a large set of periods with almost no consequences attached, whose real function is to supply courts and litigants with a statutory measure of delay; they will matter in the hands of High Courts willing to use them and will be decorative where courts are not. The third layer is the short list of provisions that move power with consequences attached: police custody obtainable in parts across forty or sixty days, a preliminary enquiry gate across the middle band of offences, deemed sanction after one hundred and twenty days of silence, trial of an absconding proclaimed offender in his absence, and an undertrial benefit withdrawn where multiple cases pend. On those five provisions the Sanhita is not a restatement of anything. It is a new settlement of the balance between investigation and liberty, and it is on those five provisions, not on the timelines, that the code will be fought over and finally judged.

Beneath the verdict sits a pattern in the code’s drafting that explains both its strengths and its limits, and it can be stated as a distribution of discretion. At the three points where the Sanhita is most praised, discretion was removed from the actor who had abused it: the station officer lost the power to refuse a first information report on territorial grounds, the government lost the power to keep a sanction request unborn, and the trial court lost the open-ended adjournment. At the three points where the Sanhita is most exposed, discretion was preserved or created: the Magistrate’s standard for custody sought in parts is unstated, the State Government’s date for the forensic duty is its own, and the electronic mode under Section 530 is a permission whose exercise depends on rules and infrastructure the code does not supply. A code is a machine for allocating discretion, and the Sanhita allocates it with a consistent bias, withdrawing it from the visible front of the process, where refusal and delay are documented and litigated daily, and leaving it in the system’s deeper rooms, the remand hearing, the notification file and the courtroom’s equipment, where the record is thinner. That bias, more than any single provision, is what the next generation of litigation under the code will be about.

What would settle the argument is a short list of observable events, and naming them is the useful way to close. The first is notification: the dates on which State Governments bring the forensic duty into force, and for which offences, will show whether Section 176(3) is a reform in progress or a reform in suspense. The second is the custody standard: a decision of the Supreme Court, whether in the pending reference on the old code or in a challenge under the Sanhita, that states what justification a Magistrate must demand for custody sought in parts will determine whether Section 187 operates as supervised flexibility or as an entitlement of the investigating agency. The third is the fate of Section 479(2), whose multiple-cases bar is the provision most exposed to an equality and liberty challenge and whose defence requires the State to justify excluding the most encumbered undertrials from a benefit enacted for undertrials. The fourth is quieter: whether High Courts, exercising superintendence over their district judiciaries, take up the consequence-free periods in the timelines table as standards for case management, because that, and not any sanction inside the code, is the mechanism by which a directory period becomes a working one. Until those events accumulate, the accurate description of the Bharatiya Nagarik Suraksha Sanhita is the one this article has defended throughout: a familiar code, re-enacted with its doctrine intact, carrying a clock that mostly does not strike, a small number of genuine transfers of power, and one transfer, custody in parts, whose direction is unmistakable. The repeal was real. The replacement is real in the places where its words moved, and the words that moved are now the ones to watch.

Frequently Asked Questions

Q: What does the Bharatiya Nagarik Suraksha Sanhita govern?

The Bharatiya Nagarik Suraksha Sanhita, 2023 is India’s code of criminal procedure. It governs how a criminal case moves from the first information given to the police through investigation, arrest and custody, the filing of the police report, cognizance, committal, the framing of charges, the conduct of the trial, the pronouncement of judgment and the machinery of appeal, revision and bail. It replaced the Code of Criminal Procedure, 1973 on 1 July 2024 and runs to 531 sections. It does not define offences or punishments, which belong to the Bharatiya Nyaya Sanhita; it supplies the process by which those offences are investigated and tried, and it binds police officers, Magistrates, Courts of Session and the High Courts in the conduct of that process.

Q: What is a zero FIR and is it now compulsory?

A zero first information report is an information about a cognizable offence recorded by a police station even though the offence was committed outside that station’s jurisdiction, given a number beginning with zero and then transferred to the station that has jurisdiction. Section 173 of the Sanhita provides that information relating to a cognizable offence may be given to the officer in charge of a police station irrespective of the area where the offence was committed, which makes the zero FIR a statutory duty rather than a matter of practice. The station cannot refuse to record the information on the ground that the incident happened elsewhere. The practice existed before the Sanhita under judicial directions and executive instructions, and the section’s contribution is to give the informant a statutory text that the receiving officer is bound by.

Q: What investigation timelines does the new procedure code impose?

The Sanhita times the investigation at several points. A preliminary enquiry under Section 173(3), where one is held, must finish within fourteen days. Investigation of the listed sexual offences under Section 193(2) must be completed within two months of the recording of information. The progress of every investigation must be communicated to the informant or the victim within ninety days under Section 193(3). A medical examination report in a rape investigation must reach the investigating officer within seven days. The custody clock in Section 187 gives the agency sixty or ninety days, depending on the offence, before the accused becomes entitled to default bail, and further investigation during a trial is to be completed within ninety days, a period the court may extend. Most of these periods state no consequence for breach.

Q: How did police custody rules change under the new code?

The total amount of police custody is unchanged: a Magistrate may authorise fifteen days of police custody in all. What changed is when those fifteen days may be taken. Under the Code of Criminal Procedure as read in Anupam J Kulkarni, police custody could be granted only within the first fifteen days of remand, and after that window the accused could be held only in judicial custody. Section 187(2) of the Sanhita allows the fifteen days to be authorised in the whole or in parts at any time during the initial forty days of the detention period where the outer limit is sixty days, or during the initial sixty days where the outer limit is ninety days. Police may therefore return for custody in instalments weeks into the investigation, which is the change practitioners contest most.

Q: Is forensic examination now mandatory for serious offences?

Section 176(3) makes forensic examination mandatory in form and phased in operation. Where information relates to an offence punishable with seven years or more, the officer in charge of the police station must cause a forensic expert to visit the scene of the offence to collect forensic evidence, and must have the process videographed on a mobile phone or another electronic device. The duty operates from a date the State Government notifies within a period of five years, and where the forensic facility for an offence is not available in the State, the State Government must notify the use of a facility in another State until its own is developed. The threshold and the duty are therefore in the statute, but the obligation switches on State by State as forensic capacity is built.

Q: Does the Bharatiya Nagarik Suraksha Sanhita allow trial in absentia?

Yes, in one confined situation. Section 356 permits a court to try a person in his absence where he has been declared a proclaimed offender under Section 84, has absconded to evade trial and there is no immediate prospect of arresting him; the section treats his absence as a waiver of the right to be present and allows judgment to be pronounced with the same effect as if he had been present. The safeguards are procedural: the trial cannot begin until ninety days have lapsed from the framing of the charge, two consecutive warrants of arrest must have issued at least thirty days apart, notice must be published in a newspaper and affixed at his last known address, a relative or friend must be informed, and an unrepresented offender must be given an advocate at the expense of the State.

Q: How has bail changed under the new procedure code?

The structure of bail is carried over from the old code with renumbered provisions: bail in non-bailable offences under Section 480, anticipatory bail under Section 482, the wider powers of the High Court and the Court of Session under Section 483, and default bail under Section 187(3) where the investigation outruns its sixty or ninety days. The substantive changes sit in the undertrial provision, Section 479. A first-time offender is now entitled to release on bond after undergoing one-third of the maximum imprisonment for the offence, the jail superintendent must apply for an undertrial’s release when the qualifying period completes, and Section 479(2) withdraws the benefit altogether where investigation, inquiry or trial in more than one offence or in multiple cases is pending against the person.

Q: How does the Bharatiya Nagarik Suraksha Sanhita treat sanction for prosecuting public servants?

Section 218 continues the rule of Section 197 of the old code: a court may not take cognizance of an offence alleged against a Judge, Magistrate or public servant for an act done in the discharge of official duty without the previous sanction of the competent government. The Sanhita adds a clock and a consequence. The government must take a decision on the request for sanction within one hundred and twenty days of receiving it, and if it fails to do so, the sanction is deemed to have been accorded. Under the old code there was no such deeming provision, and the Supreme Court confirmed in Suneeti Toteja v State of Uttar Pradesh that governmental delay, however long, could not manufacture a sanction the statute had not granted.

Q: What is an e-FIR and how can information be given electronically under the BNSS?

An e-FIR is the first information about a cognizable offence given by electronic communication rather than orally or in writing at the police station. Section 173 recognises this mode alongside oral information: the information may be sent electronically to the officer in charge, and it is taken on record once the person giving it signs it within three days. The signature requirement is the part of the mechanism most often missed; electronic information starts the process but does not complete the informant’s attestation, and the three days are the window for that step. Once the information is recorded, the informant or victim is entitled to a copy of it free of cost under Section 173(2), supplied forthwith, as with an information recorded in person.

Q: What is default bail under the Bharatiya Nagarik Suraksha Sanhita?

Default bail is the accused’s entitlement to be released when the investigating agency misses its statutory deadline for completing the investigation. Section 187(3) fixes the outer period of custody during investigation at ninety days where the offence is punishable with death, imprisonment for life or a term of ten years or more, and sixty days for any other offence. If the police report has not been filed within that period, the accused who is prepared to furnish bail must be released on bail, and the entitlement is treated in the case law under the predecessor provision as indefeasible once it has accrued and been claimed. It is the one investigation-stage timeline in the Sanhita that carries its own consequence, which is why it has done more practical work than any other period in the code.

Q: What happens if the police miss an investigation deadline under the BNSS?

For most investigation deadlines in the Sanhita, the statute states no consequence, and a missed period does not end the investigation or void its product. The two-month period for the listed sexual offences, the fourteen-day preliminary enquiry period and the ninety-day duty to update the victim are all silent about their own breach, and on the pattern of the old code such periods are likely to be read as directory: they supply a standard against which delay can be challenged before a Magistrate or a High Court, and they can support an argument about the fairness of the process, but the investigation continues. The exception is the custody clock in Section 187(3), where expiry of the sixty or ninety days matures the accused’s entitlement to default bail. The consequence, where the drafters wanted one, was written in; where it was not written in, it does not follow automatically.

Q: Is audio-video recording of search and seizure compulsory under the BNSS?

Yes. Section 105 requires that the process of search and seizure, including the preparation of the list of things seized and signed by the witnesses, be recorded through audio-video electronic means, preferably a mobile phone. The recording must be forwarded without delay to the District Magistrate or the Sub-divisional Magistrate. The old code required a signed seizure list and respectable witnesses but no recording, so the section adds a verifiable record of what was found, where, and in whose presence, which is the point on which planting and recovery disputes in trial usually turn. Whether a seizure unsupported by the required recording will be excluded, discounted or merely criticised is a question the courts will have to settle case by case, since the section, like most of the Sanhita’s process duties, does not state the consequence of non-compliance.

Q: What does the BNSS say about handcuffing a person who has been arrested?

Section 43(3) addresses handcuffing expressly for the first time in the procedure code. Keeping in view the nature and gravity of the offence, a police officer may use handcuffs while arresting a person or while producing that person before the court where the person is a habitual or repeat offender, has escaped from custody, or has committed an offence of organised crime, a terrorist act, drug-related crime, illegal possession of arms and ammunition, murder, rape, acid attack, counterfeiting of coins and currency notes, human trafficking, a sexual offence against children, or an offence against the State. The provision converts what had been governed by Supreme Court directions treating routine handcuffing as an affront to dignity into a statutory permission for the listed categories, and how far the permission displaces the older case law is one of the code’s open questions.

Q: How does the BNSS change the maximum period an undertrial prisoner can be detained?

Section 479 carries forward the rule that an undertrial who has undergone detention for one-half of the maximum imprisonment specified for the offence must be released on bail, and it changes the rule in three ways. A first-time offender, never previously convicted, qualifies earlier, on bond, at one-third of the maximum period. Offences punishable with life imprisonment are excluded alongside offences punishable with death, where the old code excluded death alone. And Section 479(2) provides that where an investigation, inquiry or trial in more than one offence or in multiple cases is pending against the person, the release under the section is not available at all. The jail superintendent is placed under a duty to apply to the court for the prisoner’s release when the half or the third, as the case may be, is completed.

Q: Within what period must a charge be framed under the BNSS?

The Sanhita fixes sixty days from the first hearing on charge. In a trial before a Court of Session the charge must be framed within that period under Section 251, and the same sixty-day period from the first hearing on charge applies to warrant cases tried by a Magistrate under Section 263. The periods around the charge move with it: an accused in a sessions case may apply for discharge within sixty days of the commitment of the case under Section 250, and an accused in a warrant case within sixty days of the supply of documents under Section 262. None of these provisions states what follows if the sixty days pass without a charge, so the periods operate as the court’s own statutory timetable, breach of which is a ground for complaint and superintendence rather than a defect that dissolves the case.

Q: How quickly must a court pronounce judgment after a trial under the BNSS?

Two provisions time the judgment. In a sessions trial, Section 258 requires the Judge to give judgment as soon as possible and within thirty days of the completion of arguments, a period that may be extended to forty-five days for reasons recorded in writing. Section 392 sets the general outer limit for every court of original jurisdiction: judgment is to be pronounced within forty-five days of the termination of the trial. Under the old code there was no general statutory period for delivering judgment, and reserved judgments that waited for months were a standing complaint addressed only by appellate exhortation. The Sanhita’s periods state no consequence for a late judgment, so their force is disciplinary and appellate rather than automatic, but they convert a grievance about delay into the breach of a written standard.

Q: Must the police tell the victim how the investigation is progressing under the BNSS?

Yes. Section 193(3) requires the police officer to inform the informant or the victim of the progress of the investigation within ninety days, and the information may be given by any means including electronic communication. The same provision requires the officer to communicate the action taken to the person who first gave the information about the offence. Under the old code the informant’s statutory entitlements effectively ended with the free copy of the first information report, and a complainant who wanted to know what had become of the case depended on the station’s goodwill or on approaching the Magistrate. The ninety-day duty is one of the Sanhita’s clearer instances of a new obligation in substance, although, in keeping with most of the code’s process duties, it states no consequence if the ninety days pass in silence.

Q: Can trials and other proceedings be held through video conferencing under the BNSS?

Yes. Section 530 provides that all trials, inquiries and proceedings under the Sanhita may be held in electronic mode, by the use of electronic communication or audio-video electronic means, and it lists the stages expressly: the issuance, service and execution of summons and warrants, the examination of the complainant and witnesses, the recording of evidence in inquiries and trials, and all appellate or other proceedings. The old code contained scattered provisions for particular electronic steps, built up by amendment and by court practice directions; Section 530 generalises them into a single enabling provision covering the whole life of a case. The provision permits rather than compels, so the extent of electronic trial in a given district depends on the rules, infrastructure and practice directions of the courts there.

Q: What special safeguards does the BNSS give victims of sexual offences during investigation?

The investigation of sexual offences carries its own cluster of duties. Investigation of the offences listed in Section 193(2), including the principal rape offences under the Bharatiya Nyaya Sanhita and specified offences under the Protection of Children from Sexual Offences Act, must be completed within two months of the recording of information. The medical examination of the victim is to be conducted by a registered medical practitioner with consent, and Section 184(6) requires the practitioner to forward the report to the investigating officer within seven days. The victim’s statement is recorded with the safeguards the old code had built up, including recording by a woman police officer in the settings the provisions specify, and the trial of the listed offences is itself timed by Section 346 at two months from the filing of the chargesheet. The pattern is of duties distributed across the case rather than gathered in one place.

Q: What changed for summary trials under the Bharatiya Nagarik Suraksha Sanhita?

Section 283 widens the summary trial, the shortened procedure for minor offences. A Chief Judicial Magistrate or a Magistrate of the first class must try the listed offences summarily, and may also, after giving the accused a reasonable opportunity of being heard and recording reasons in writing, try in a summary way any offence not punishable with death, imprisonment for life or imprisonment for a term exceeding three years. The old code drew that line at two years, so the Sanhita moves a band of offences into the faster track. The property-value ceiling for the listed theft and property offences that must be tried summarily rises from two hundred rupees under the old code to twenty thousand rupees, a change that restores to the provision a field of operation that the old figure, unrevised across decades, had reduced to a formality.