INTRODUCTION
The criminal justice system rests on the presumption of innocence, yet an accused person may spend months or years in custody before a trial concludes. This tension makes bail one of the most consequential, and most contested, areas of Indian criminal procedure.
Bail is frequently treated as a procedural formality, but it engages the constitutional guarantee of personal liberty under Article 21. Prolonged pre-trial custody risks converting an interim procedural measure into a substantive punishment imposed before guilt is established.
The Supreme Court has repeatedly affirmed that ‘bail is the rule and jail is the exception’, reflecting the view that pre-conviction imprisonment should not function as punishment.[1] This does not mean release is automatic: courts weigh the gravity of the allegation against the risk of absconding, witness tampering or evidence interference. The Bharatiya Nagarik Suraksha Sanhita 2023 (BNSS), replacing the Code of Criminal Procedure 1973 (CrPC), has altered specific elements of the bail framework, most significantly through a new detention ceiling in section 479, and invites renewed scrutiny of whether the liberty principle is applied consistently, or whether it remains, as this article argues, a doctrine more secure in judgment than in outcome.[2]
CONSTITUTIONAL FOUNDATION
India has no express fundamental right to bail, but Article 21 supplies its basis: any procedure depriving a person of liberty must be fair, just and reasonable.[3] In Gudikanti Narasimhulu v Public Prosecutor, the Court situated bail within the accused’s social and economic circumstances, recognising that poverty should never operate as a silent ground for continued detention.[4] This shifts bail jurisprudence from an offence-centric inquiry toward one attentive to structural disadvantage, a shift the BNSS’s drafting does not obviously carry forward.
BAIL UNDER THE BNSS: WHAT HAS ACTUALLY CHANGED
The BNSS retains the CrPC’s bailable/non-bailable structure but alters it in identifiable respects. Section 478 (formerly section 436) governs bailable offences and now expressly extends bail to an indigent accused unable to furnish sureties within a week of arrest. Section 480 (formerly section 437) governs non-bailable offences, adding an explicit bar on bail where the accusation carries death or life imprisonment and the court finds reasonable grounds to believe it true. Section 483 (formerly section 439) preserves the wider powers of Sessions Courts and High Courts.[5]
The most consequential change is section 479, replacing section 436A CrPC. Section 436A applied a uniform one-half ceiling to all undertrials and imposed no duty on prison authorities to act unprompted. Section 479 innovates in three respects: it introduces a lower one-third threshold for first-time offenders, unknown to section 436A; section 479(3) imposes an affirmative duty on the jail superintendent to apply for release once a threshold is reached, rather than leaving the initiative to the accused; and section 479(2) bars release where the undertrial faces proceedings in more than one case — a net narrowing absent from section 436A.[6]
SECTION 479: CONDITIONS AND EXCEPTIONS
The provision is not an entirely automatic ceiling. Four qualifications limit it. First, section 479(1) excludes any offence carrying death or life imprisonment absolutely, regardless of time served — stricter than section 436A, under which even life-imprisonment cases could in principle be considered. Second, section 479(2) denies release wherever proceedings in more than one case are pending, irrespective of individual eligibility. Third, courts retain discretion under the second proviso to continue detention beyond the threshold where reasons are recorded in writing after hearing the public prosecutor, so the ceiling is not judge-proof. Fourth, the explanation excludes delay attributable to the accused from the computation, so dilatory defence conduct does not accelerate eligibility. Section 479 thus functions less as a self-executing cap than as a conditional entitlement whose activation depends on offence-type, absence of parallel proceedings, and institutional follow-through by the jail superintendent and the district Undertrial Review Committee (UTRC).[7]
JUDICIAL RECOGNITION OF LIBERTY
In Moti Ram v State of Madhya Pradesh, the Court cautioned against onerous monetary bail conditions, recognising that a formally available right to bail is meaningless if unaffordable.[8] Two accused persons facing identical charges may experience the law differently — one released, one detained — not through any legally relevant distinction but through unequal capacity to furnish a surety. In Sanjay Chandra v CBI, the Court held that bail is neither punitive nor preventive, and that the accused’s likely availability for trial is the central inquiry.[9]
DISTINGUISHING ANTIL AND SISODIA
Satender Kumar Antil and Manish Sisodia are often cited together as evidence of a consolidated liberty jurisprudence, but they operate at different stages and should not be conflated. Antil addresses arrest and the pre-chargesheet stage: it requires compliance with sections 41 and 41A CrPC before arrest, and classifies offences into categories — broadly, offences punishable with seven years or less, offences punishable with more or with death or life imprisonment, and offences under special statutes with stringent bail provisions such as the NDPS Act and PMLA — prescribing that ordinary summons, not arrest, should generally follow chargesheet filing in the least serious category.[10] Its concern is preventing unnecessary custody from arising at all.
Sisodia, by contrast, addresses the consequences of incarceration that has already occurred, in a case governed by the stringent twin conditions of section 45 of the Prevention of Money Laundering Act 2002. The Court held that the Article 21 right to a speedy trial must be read into section 45 itself, so that indefinite incarceration during an uncertain trial timeline becomes an independent ground for bail, displacing the ordinarily higher threshold special statutes impose, where delay is not attributable to the accused.[11]
THE GAP BETWEEN LAW AND PRACTICE
Implementation lags doctrine, and recent official data substantiate the point. The National Crime Records Bureau’s Prison Statistics India 2024 report records undertrials at approximately 73% of India’s prison population at end-2024 — 371,440 of roughly 511,000 prisoners — down from a pandemic-era peak of 77% in 2021 but well above the 66% recorded in 2005.[12] The India Justice Report 2025 similarly recorded undertrials at 76% of the custodial population, with 301 prisons holding no convicted person at all.[13] Two in three undertrials belong to Scheduled Caste, Scheduled Tribe or Other Backward Class communities, lending empirical weight to Gudikanti Narasimhulu‘s concern that structural disadvantage, not legal entitlement, often determines who remains in custody.[14]
This gap is institutional rather than doctrinal, and the Court’s own monitoring in In Re: Inhuman Conditions in 1382 Prisons illustrates its scale. Having directed retrospective application of section 479 in August 2024, the Court subsequently found that fewer than twenty of thirty-six states and union territories had filed compliance reports, and that four states had not responded at all to a follow-up order.[15] Separate monitoring of UTRCs across 357 districts found only a small fraction fully complied with the quarterly review mandate.[16] A right dependent on jail superintendents and district committees acting unprompted, against caseload pressure, is structurally weaker in operation than its language suggests.
IS THE PRINCIPLE REALLY BEING APPLIED?
Partially. The doctrine is settled: personal liberty, the limited purpose of pre-trial detention, and a conditional statutory ceiling all point the same way. But formal entitlement and practical liberty diverge whenever delay, cost, offence classification or administrative inattention intervene. ‘Bail is the rule’ cannot mean an unconditional right to release; courts must still protect witnesses, evidence and investigative integrity. The more useful question is not whether the principle should persist, but whether the system supplies the institutional conditions needed to make it self-executing rather than dependent on an accused’s ability to assert it.
THE WAY FORWARD
Four measures would narrow the gap, each with some institutional or judicial basis already. First, arrest should follow the necessity assessment Antil prescribes, rather than automatically accompanying registration of an offence. Second, bail conditions should be calibrated to actual means, consistent with Moti Ram. Third, trial delay itself needs institutional attention, since bail cannot resolve a crisis rooted in slow adjudication, as Sisodia recognises. Fourth, section 479 should be administered through proactive, systematic tracking rather than individual application — not a novel proposal, since the Supreme Court has already directed UTRCs, jail superintendents and Legal Services Authorities to treat eligibility identification as continuous, as an undertrial may cross the threshold the day after data was last collected.[17] The Ministry of Home Affairs has separately written to all states reiterating the section 479(3) duty on superintendents.[18] The uneven compliance recorded in the Court’s own monitoring shows precisely where integrated tracking — for instance through the e-Prisons database recommended to state governments — would close the gap between entitlement and release.[19]
CONCLUSION
‘Bail is the rule and jail is the exception’ remains a foundational protection of personal liberty, one the judiciary continues to develop, including through the BNSS’s new but conditional detention ceiling. Yet undertrial detention data, uneven UTRC compliance and unaffordable bail conditions show that formal strength has not been matched by institutional delivery. The task ahead is not to choose between liberty and public safety, but to build the administrative machinery — proactive detention tracking, proportionate conditions, faster trials — that allows a settled legal principle to operate as lived reality rather than a phrase repeated in judgments.
Author(s) Name: Niharika Aggarwal
References:
[1]State of Rajasthan v Balchand (1977) 4 SCC 308.
[2]Bharatiya Nagarik Suraksha Sanhita 2023, Act 46 of 2023.
[3]Constitution of India 1950, art 21.
[4]Gudikanti Narasimhulu v Public Prosecutor (1978) 1 SCC 240.
[5]Bharatiya Nagarik Suraksha Sanhita 2023, ss 478, 480, 483; cf Code of Criminal Procedure 1973, ss 436, 437, 439.
[6]Bharatiya Nagarik Suraksha Sanhita 2023, s 479; Code of Criminal Procedure 1973, s 436A.
[7]Bharatiya Nagarik Suraksha Sanhita 2023, s 479(1)-(3) and provisos.
[8]Moti Ram v State of Madhya Pradesh (1978) 4 SCC 47.
[9]Sanjay Chandra v Central Bureau of Investigation (2012) 1 SCC 40.
[10]Satender Kumar Antil v Central Bureau of Investigation (2022) 10 SCC 51.
[11]Manish Sisodia v Directorate of Enforcement 2024 SCC OnLine SC 1920.
[12]National Crime Records Bureau, Prison Statistics India 2024 (Ministry of Home Affairs 2026).
[13]Tata Trusts, India Justice Report 2025 (Tata Trusts 2025).
[14]National Crime Records Bureau (n 12); ‘Nearly 3 in 4 Indians in Prison Are Under Trial’ The News Minute (Pune, 25 June 2026).
[15]In Re: Inhuman Conditions in 1382 Prisons, Writ Petition (Civil) No 406 of 2013, Order of the Supreme Court of India dated 23 August 2024; see also ‘Justice Must Extend to the Last Person: SC Calls for Continuous Reporting on Implementation of Section 479 of BNSS’ (Verdictum, 2025) < https://www.verdictum.in/court-updates/supreme-court/in-re-inhuman-conditions-in-1382-prisons-release-undertrial-prisoners-prolonged-incarceration-1559091 >accessed 9 September 2026.
[16]Commonwealth Human Rights Initiative, ‘Presentation on the National Report on Under Trial Review Committees’ (2017) accessed 9 September 2026.
[17]‘Take Proactive Steps to Release Deserving Undertrial Prisoners Under S.479 BNSS: Supreme Court to States/UTs’ (LiveLaw, 2024) < https://www.livelaw.in/top-stories/take-proactive-steps-to-release-deserving-undertrial-prisoners-under-s479-bnss-supreme-court-to-statesuts-273246 > accessed 9 September 2026.
[18]‘Centre Asks States to Implement BNSS Section 479 to Ensure Relief to Undertrial Prisoners’ Deccan Herald (New Delhi) < https://www.deccanherald.com/india/centre-asks-states-to-implement-bnss-section-479-to-ensure-relief-to-undertrial-prisoners-3518612 > accessed 9 September 2026.
[19]‘SC Fine Tunes Under Trial Review Committees’ (LiveLaw) < https://www.livelaw.in/sc-fine-tunes-under-trial-review-committees-read-order > accessed 9 September 2026.