No part of India's matrimonial criminal law has generated more sustained controversy than the cluster of provisions built around dowry. The penal scaffolding is well known: Section 3 and Section 4 of the Dowry Prohibition Act, 1961, reinforced by Section 498A of the Indian Penal Code (cruelty by husband or relatives) and Section 304B IPC (dowry death). Together they were meant to be a shield for vulnerable wives. Yet for two decades judges, law commissions and committees have asked whether the shield has, in too many cases, been turned into an ‘assassin's weapon’. This chapter maps the misuse-and-reform debate in full: the ‘legal terrorism’ rhetoric, the arrest-restraint jurisprudence of Arnesh Kumar, the rise and fall of Family Welfare Committees, the over-implication of relatives, the data wars over conviction rates, and the legislative responses that culminated in the Bharatiya Nyaya Sanhita, 2023.

The penal architecture under debate

To assess the misuse argument one must first be precise about which provisions are in play, because the ‘dowry law’ is in truth a network of statutes rather than a single section. The Dowry Prohibition Act, 1961 supplies the substantive prohibitions: Section 3 punishes the giving or taking of dowry with imprisonment of not less than five years and a fine of not less than fifteen thousand rupees or the amount of the dowry, whichever is more; Section 4 punishes the mere demand for dowry with imprisonment of not less than six months extending to two years and fine up to ten thousand rupees. Section 7 makes these offences cognizable for limited investigative purposes, non-bailable and non-compoundable, and Section 8A places the burden of proving that an article was not given as dowry on the person who took it. These are buttressed by Section 8B, which authorises State Governments to appoint Dowry Prohibition Officers.

The provisions that actually generate the bulk of litigation, however, sit in the penal code rather than the 1961 Act. Section 498A IPC criminalises ‘cruelty’ by a husband or his relatives, including harassment to coerce an unlawful demand for property — in substance a dowry demand. Section 304B IPC, read with the rebuttable presumption in Section 113B of the Evidence Act, treats the unnatural death of a woman within seven years of marriage, preceded by dowry-linked cruelty ‘soon before her death’, as a ‘dowry death’ punishable with a minimum of seven years and up to life imprisonment. It is this overlapping, cognizable, non-bailable cluster — not the comparatively mild Section 4 — that the reform debate principally targets. Readers should keep that distinction in mind: most ‘dowry law misuse’ complaints are really complaints about Section 498A IPC.

Origin of the misuse narrative: from protection to backlash

Section 498A was inserted by the Criminal Law (Second Amendment) Act, 1983, in direct response to a wave of bride-burnings and dowry suicides that had shocked public opinion in the late 1970s. Parliament deliberately made the offence cognizable and non-bailable so that police could act swiftly and the accused could not secure routine release. The same statutory package introduced Section 113A of the Evidence Act (presumption as to abetment of suicide) and was followed in 1986 by Section 304B IPC and Section 113B of the Evidence Act. The architecture was, by design, pro-complainant: it lowered the threshold for state intervention because the legislature believed under-enforcement, not over-enforcement, was the real evil.

Almost as soon as the provision matured, a counter-narrative emerged. Trial courts and High Courts began to remark that the very features which made Section 498A effective — immediate arrest, non-bailability, social stigma — also made it a powerful instrument of pressure in marital breakdowns that had nothing to do with dowry. By the late 1990s the refrain that the section was being used to settle scores, extract divorce settlements, or simply to coerce a reluctant husband had become a fixture of matrimonial litigation. The debate thus pits two genuine concerns against each other: the historical and continuing reality of dowry violence on one side, and the danger of a criminal process weaponised in family disputes on the other. This is the framing to which the Dowry Prohibition Act hub repeatedly returns.

The phrase that crystallised the misuse debate came from the Supreme Court itself. In Sushil Kumar Sharma v. Union of India, (2005) 6 SCC 281, the petitioner challenged the constitutional validity of Section 498A IPC, arguing that its potential for abuse rendered it ultra vires. The Court, per Pasayat J., declined to strike the provision down — mere possibility of misuse, it held, is no ground to declare a law unconstitutional, since the remedy lies in curbing the abuse, not abolishing the safeguard. But in the course of upholding the section the Bench delivered the line that has been quoted in virtually every subsequent misuse case: ‘by misuse of the provision a new legal terrorism can be unleashed.’ The Court added the now-canonical metaphor that the provision is intended to be used as a ‘shield and not an assassin's weapon’.

The significance of Sushil Kumar Sharma is twofold. First, it settled that the response to misuse is judicial and procedural vigilance, not invalidation — a holding that anchors all later reform attempts within the existing statute. Second, by judicially endorsing the ‘legal terrorism’ framing, the Court lent its enormous authority to the misuse narrative, which thereafter could no longer be dismissed as mere men's-rights rhetoric. Every committee report and every guideline judgment that followed traces its starting premise to this case.

The over-implication of relatives

If there is one factual pattern that unites the misuse cases, it is the indiscriminate naming of the husband's extended family. The Court flagged this long before the modern guideline era. In Kans Raj v. State of Punjab, (2000) 5 SCC 207, a dowry-death appeal, the Court warned that ‘a tendency has, however, developed for roping in all relations of the in-laws of the deceased in the matters of dowry deaths’, and cautioned that such over-enthusiasm to convict the maximum number of persons in fact weakens the prosecution even against the genuinely guilty. Only the husband's conviction was sustained.

This concern hardened into a quashing principle in Preeti Gupta v. State of Jharkhand, (2010) 7 SCC 667, where the Court observed that ‘it is a matter of common experience that most of these complaints under Section 498A IPC are filed in the heat of the moment over trivial issues’ and that ‘exaggerated versions of the incident’ and ‘over implication’ of relatives who often live in different cities are routine. The Court urged the legislature to take a serious relook. Two years later, in Geeta Mehrotra v. State of U.P., (2012) 10 SCC 741, it quashed proceedings against the husband's brother and sister where the FIR contained only a ‘casual reference’ to their names without specific allegations, holding that distant relatives cannot be dragged through trial on general, omnibus averments. These cases supply the doctrinal backbone for thousands of High Court quashing orders under Section 482 CrPC.

Arnesh Kumar: restraining the reflex arrest

The single most consequential intervention is Arnesh Kumar v. State of Bihar, (2014) 8 SCC 273. The appellant, apprehending arrest in a Section 498A case coupled with Section 4 of the Dowry Prohibition Act, moved the Supreme Court. The Bench (Chandramauli Kr. Prasad and Pinaki Chandra Ghose JJ.) used the occasion to address what it described as the casual, mechanical manner in which police arrested husbands and their relatives the moment an FIR was lodged. It noted that Section 498A is a cognizable and non-bailable offence carrying punishment up to three years, and that it had ‘acquired the dubious distinction of being one of the most abused sections’.

The Court issued binding directions anchored in Section 41(1)(b) and Section 41A CrPC: in offences punishable with up to seven years' imprisonment, police must not arrest automatically but must first satisfy themselves, on a reasoned checklist, that arrest is genuinely necessary, and must furnish that reasoning to the Magistrate, who in turn must apply mind before authorising detention. Non-compliance was made punishable as contempt and departmental delinquency. Arnesh Kumar thus converted the misuse concern into an enforceable arrest-screening protocol. Crucially, it did not dilute the substantive offence at all — a point reformers on the complainant side stress — but it dramatically changed day-to-day enforcement, and remains the most cited matrimonial-law judgment in bail and anticipatory-bail practice.

Family Welfare Committees: the Rajesh Sharma experiment

The boldest judicial attempt to engineer a pre-arrest filter came in Rajesh Sharma v. State of U.P., (2018) 10 SCC 472 (judgment dated 27 July 2017). A two-Judge Bench (A.K. Goel and U.U. Lalit JJ.), persuaded that a large proportion of Section 498A complaints were exaggerated, directed that every district constitute one or more Family Welfare Committees through the District Legal Services Authority. No arrest was ordinarily to be made until such a Committee — composed of lay volunteers — examined the complaint and submitted a report, generally within a month. The Court also directed that bail applications be decided expeditiously and permitted dispensing with personal appearance of outstation relatives.

The directions were immediately and fiercely contested. Critics argued that interposing an extra-statutory, non-judicial body between a victim and the police rewrote the Code of Criminal Procedure by judicial fiat, delayed protection in genuine cruelty cases, and effectively diluted a cognizable offence without legislative sanction. The episode is the high-water mark of judicial activism in this field and a standard examination illustration of the tension between curbing misuse and respecting the separation of powers.

The course-correction: Social Action Forum for Manav Adhikar

The Family Welfare Committee scheme survived barely a year. In Social Action Forum for Manav Adhikar v. Union of India, (2018) 10 SCC 443, a three-Judge Bench led by Dipak Misra CJ (with A.M. Khanwilkar and D.Y. Chandrachud JJ.) reviewed the Rajesh Sharma directions and held that the constitution of Family Welfare Committees and the directions regarding their functioning were ‘impermissible’ — the judiciary could not create such bodies and graft them onto the criminal process, which was the legislature's domain. The Court accordingly recalled the Committee mechanism.

Importantly, the Bench did not abandon the anti-misuse project; it relocated it within existing law. It directed that investigating officers be suitably trained and that the safeguards already available — the arrest discipline of Arnesh Kumar, the FIR-registration rules of Lalita Kumari, and the quashing power under Section 482 CrPC — be conscientiously applied. The net effect of the Rajesh Sharma–Manav Adhikar sequence is the prevailing position: no extra-statutory committee, but real procedural vigilance against frivolous prosecution. For students, the pair is the definitive case study of the Court testing, and then withdrawing, a controversial reform.

Modern quashing jurisprudence: Achin Gupta and Dara Lakshmi Narayana

Recent benches have continued to police misuse through the quashing route while sharpening the rhetoric. In Achin Gupta v. State of Haryana, 2024 INSC 369, the Supreme Court quashed an FIR under Sections 498A, 406, 323 and 506 IPC, finding the allegations vague and evidently retaliatory to the husband's prior divorce petition. The Court went further than disposing of the case: it expressly urged Parliament to relook at Sections 85 and 86 of the then-pending Bharatiya Nyaya Sanhita, 2023 — the provisions that re-enact Section 498A IPC — before they came into force, so that the legislature might build in safeguards against abuse.

Later the same year, in Dara Lakshmi Narayana v. State of Telangana, 2024 INSC 953 (decided 10 December 2024), a Bench of B.V. Nagarathna and N. Kotiswar Singh JJ. quashed proceedings under Section 498A IPC and Sections 3 and 4 of the Dowry Prohibition Act, holding that vague, omnibus allegations levelled to wreak personal vendetta against the husband and his entire family cannot be allowed to proceed. The Court reiterated that genuine victims must be protected, but that the misuse of well-intentioned provisions to settle scores cannot be countenanced. These 2024 decisions show that, post-Manav Adhikar, the chosen instrument against misuse is individualised judicial scrutiny at the threshold rather than blanket procedural barriers.

Committees and Law Commission: the reform proposals

The institutional reform debate has run on a parallel track to the litigation. The Committee on Reforms of the Criminal Justice System chaired by Justice V.S. Malimath (2003) recommended that Section 498A be made bailable and compoundable, reasoning that its non-compoundable character forecloses reconciliation and harms estranged spouses who might otherwise settle. This recommendation was endorsed by the Department-related Parliamentary Standing Committee on Home Affairs.

The Law Commission of India returned to the question in its 243rd Report (2012), prepared specifically on Section 498A. The Commission accepted that there was misuse and that innocent relatives were being harassed, but it declined to recommend making the offence bailable, fearing this would deprive genuine victims of immediate protection. It did, however, support making the offence compoundable with the permission of the court, subject to judicial scrutiny of the settlement, and recommended better-trained conciliation and improved investigation rather than dilution of the substantive law. The persistent divergence — Malimath and men's-rights advocates pressing for compoundability and bail, the Law Commission and women's groups resisting any weakening of the protective core — is the heart of the legislative reform debate and a frequent essay topic. The object and background of the parent Act is the necessary backdrop to evaluating these proposals.

The data war: what do conviction rates prove?

Much of the misuse argument rests on statistics, and the statistics are genuinely contested. National Crime Records Bureau data consistently show that Section 498A generates a very large volume of cases — over a lakh annually — while the conviction rate hovers in the low to mid-teens, among the lowest for any IPC offence, with the bulk of tried cases ending in acquittal. Proponents of reform read this as direct proof of large-scale false complaints.

Sociologists and several commentators caution against that inference. A low conviction rate, they argue, is not the same as a high false-complaint rate. Chargesheeting rates under Section 498A are in fact high, indicating that police find prima facie material in most cases. Acquittals are driven by causes unrelated to falsity: out-of-court compromises and reconciliations (which the non-compoundable bar pushes into acquittal rather than withdrawal), hostile witnesses, complainants losing interest after settlement, weak investigation, and the ordinary evidentiary difficulty of proving private cruelty. The honest position for an examinee is that the data establish neither side conclusively: they show a provision that is heavily invoked and rarely results in conviction, but the reasons are multi-causal. This nuance is precisely what the Supreme Court was alive to even as it issued anti-misuse directions.

Dowry death: a different misuse calculus

It is analytically important to separate Section 498A from Section 304B IPC. Dowry-death prosecutions carry a minimum seven-year sentence and a statutory presumption under Section 113B of the Evidence Act, and they are invoked only where a woman has actually died unnaturally within seven years of marriage. The over-implication concern still surfaces here — as Kans Raj shows, prosecutors and bereaved families sometimes name the whole household — but the ‘trivial complaint’ critique that dominates the Section 498A debate has little purchase where there is a corpse.

Courts have therefore drawn a sharper line in 304B cases: while insisting that vague allegations against distant relatives be discarded, they have firmly upheld the presumption against the husband and immediate in-laws where the ‘soon before death’ nexus of dowry cruelty is established. The lesson is that the misuse debate is not monolithic. The concern is acute at the demand-and-cruelty end (Section 4 of the Act and Section 498A IPC), where the offence is easy to allege and hard to disprove, and far weaker at the dowry-death end, where a high evidentiary and factual threshold already exists. Conflating the two is a common analytical error.

The legislative endgame: Bharatiya Nyaya Sanhita, 2023

The reform debate has now passed into a new statutory frame. With the repeal of the Indian Penal Code, the cruelty offence has been re-enacted, in identical substance, as Sections 85 and 86 of the Bharatiya Nyaya Sanhita, 2023: Section 85 reproduces the offence and three-year-plus-fine punishment, and Section 86 reproduces the definition of ‘cruelty’. Dowry death reappears as Section 80 BNS. Notably, despite the express plea in Achin Gupta that Parliament build in anti-misuse safeguards before the new code commenced, the legislature retained the provisions without substantive change. The Dowry Prohibition Act, 1961 itself — Sections 3, 4, 4A, 6, 7 and 8B — was left untouched and continues to operate alongside the BNS.

The practical consequence is continuity rather than rupture: the substantive law of dowry cruelty is the same as it was, merely renumbered, and the entire body of misuse jurisprudence — Sushil Kumar Sharma, Arnesh Kumar, Preeti Gupta, Manav Adhikar — carries over by analogy. The arrest-restraint logic of Arnesh Kumar, now anchored in the corresponding provisions of the Bharatiya Nagarik Suraksha Sanhita, 2023, continues to govern. For the foreseeable future, then, the management of misuse will remain primarily a matter of judicial discipline at the arrest and quashing stages, not legislative redesign — an outcome that satisfies neither pole of the long-running debate.

Balancing the competing imperatives

The enduring difficulty is that both sides of the debate are correct in part. Dowry violence remains a grim social reality; under-reporting, not over-reporting, is the dominant problem in large parts of the country, and any dilution of protective provisions risks abandoning genuine victims who already struggle to secure convictions. At the same time, the deployment of a cognizable, non-bailable offence as a bargaining chip in matrimonial disputes — sweeping in elderly parents, married sisters and relatives living in distant cities — is a real and documented harm to which the courts have rightly responded.

The mature judicial answer, distilled across two decades, is calibration rather than abolition or unrestrained enforcement: keep the substantive offence intact for the deserving victim, but discipline the process so that arrest is necessity-driven, omnibus allegations are filtered at the threshold, and reconciliation is not foreclosed where the parties genuinely wish it. Whether that calibration is best achieved by guideline judgments, by making the offence compoundable, or by leaving it to case-by-case quashing remains unresolved. What is settled, after Sushil Kumar Sharma and Manav Adhikar, is that the answer must be found within the law — through procedural safeguards and conscientious adjudication — rather than by dismantling a provision born of genuine and continuing need. Readers should connect this discussion with the introduction and object of the Act and with the substantive penalties under Section 4.

Frequently asked questions

Did the Supreme Court ever declare Section 498A IPC unconstitutional for being prone to misuse?

No. In Sushil Kumar Sharma v. Union of India, (2005) 6 SCC 281, the Court refused to strike down Section 498A, holding that the mere possibility of misuse is no ground to invalidate a law. It famously warned that misuse could unleash ‘a new legal terrorism’ and that the provision must be a shield and not an assassin's weapon, but located the remedy in curbing abuse rather than abolishing the protection.

What are the Arnesh Kumar guidelines and which provision do they rest on?

In Arnesh Kumar v. State of Bihar, (2014) 8 SCC 273, the Court directed that for offences punishable with up to seven years (such as Section 498A IPC and Section 4 of the Dowry Prohibition Act), police must not arrest automatically. Relying on Sections 41(1)(b) and 41A CrPC, it required officers to record reasons showing arrest is necessary and Magistrates to apply mind before authorising detention, on pain of contempt and departmental action.

What happened to the Family Welfare Committees directed in Rajesh Sharma?

They were recalled. Rajesh Sharma v. State of U.P., (2018) 10 SCC 472, directed that complaints be screened by district Family Welfare Committees before arrest. In Social Action Forum for Manav Adhikar v. Union of India, (2018) 10 SCC 443, a larger Bench held the creation of such extra-statutory committees ‘impermissible’ and recalled them, directing instead that the existing safeguards in Arnesh Kumar and Lalita Kumari be applied.

How do courts treat the naming of distant relatives in dowry complaints?

With great caution. In Kans Raj v. State of Punjab, (2000) 5 SCC 207, the Court deprecated the ‘tendency to rope in’ all relations. In Preeti Gupta v. State of Jharkhand, (2010) 7 SCC 667, and Geeta Mehrotra v. State of U.P., (2012) 10 SCC 741, it quashed proceedings against relatives named only in vague, omnibus terms without specific allegations, holding that over-implication cannot survive scrutiny under Section 482 CrPC.

Does the low conviction rate under Section 498A prove that most complaints are false?

Not on its own. NCRB data show high case volumes and conviction rates in the low-to-mid teens, but chargesheeting rates are high, indicating prima facie material. Acquittals are driven by out-of-court settlements (forced into acquittal by the non-compoundable bar), reconciliation, hostile witnesses and weak investigation, not solely by falsity. The data therefore support neither the misuse claim nor its denial conclusively.

Has the cruelty offence changed under the Bharatiya Nyaya Sanhita, 2023?

Not in substance. Section 498A IPC is re-enacted as Sections 85 and 86 BNS, and dowry death as Section 80 BNS, with the same ingredients and punishment. Despite the plea in Achin Gupta v. State of Haryana, 2024 INSC 369, to add safeguards before commencement, Parliament retained the provisions unchanged, and the Dowry Prohibition Act, 1961 continues to operate alongside the new code.