Khushal Rao v. State of Bombay: a dying declaration as the sole basis of conviction
In 1957 a three-judge bench held that no rule requires a dying declaration to be corroborated; a true, voluntary declaration can by itself sustain a conviction.
- Court
- Supreme Court of India
- Citation
- AIR 1958 SC 22
- Bench
- B.P. Sinha, J., S.K. Das, J., J.L. Kapur, J.
- Decided
- 25 September 1957
The facts in brief
The deceased, Baboolal, was attacked in a lane in Nagpur on 12 February 1956. Grievously injured, he made not one but three successive dying declarations before he died — to a doctor, to a Sub-Inspector of police, and to a Magistrate. The declarations named his assailant.
On the strength of those declarations the accused, Khushal Rao, was convicted of murder under Section 302 of the Indian Penal Code. The conviction travelled to the Supreme Court, where the central question was a matter of principle as much as fact: could a conviction for murder rest on dying declarations alone, without independent corroboration, and what weight should attach to a declaration recorded by a magistrate in question-and-answer form?
The evidentiary question
A dying declaration is admitted under Section 32(1) of the Evidence Act as an exception to the rule against hearsay, on the rationale that a person facing imminent death has no motive to lie. But its peculiar character — the maker is gone and cannot be cross-examined — had bred a persistent unease. Some courts had treated a dying declaration as a "weak" piece of evidence that always required corroboration before it could be acted upon, by analogy to the rule of prudence governing accomplice testimony.
The question for the three-judge bench was whether that unease had hardened into a rule. Must a dying declaration always be corroborated, or can a court convict on a declaration it finds true and voluntary, standing alone?
The stakes of the question were practical as well as theoretical. Dying declarations arise in the gravest cases — homicides, dowry deaths, acid attacks — and frequently they are the best evidence available, because the one person who knew the assailant has named him with his last breath. A rigid rule of mandatory corroboration would, in many such cases, mean acquittal not because the declaration was doubted but because no independent witness happened to be present. Yet the opposite danger was equally real: a declaration is uniquely difficult to test, and a court that accepts one uncritically forfeits the safeguard of cross-examination. The bench had to find the line between treating a dying declaration as inherently suspect and treating it as immune from scrutiny.
What the Court held
Sinha, J., delivering the judgment of the bench, rejected any inflexible rule of corroboration.
A dying declaration recorded by a competent magistrate in the proper manner, that is to say, in the form of questions and answers, and, as far as practicable, in the words of the maker of the declaration, stands on a much higher footing.
The Court held that there is neither a rule of law nor a rule of prudence to the effect that a dying declaration cannot be acted upon without corroboration. A dying declaration is not, as a class, weaker than other evidence; it stands on the same footing as any other piece of evidence and has to be judged in the light of the surrounding circumstances and with reference to the principles governing the weighing of evidence. If the court, on a careful scrutiny, is satisfied that the declaration is true and voluntary, it can base a conviction on it without corroboration.
The bench drew a gradation among declarations by their mode of recording. A declaration recorded by a competent magistrate in question-and-answer form, and as far as practicable in the declarant's own words, stands on a higher footing than one depending on the oral testimony of bystanders, because the magistrate's record reduces the risk of error, omission, or embellishment. On the facts, with three concordant declarations including one recorded by a Magistrate, the conviction under Section 302 was upheld.
The doctrinal architecture
Khushal Rao is the foundational Indian authority on the sufficiency of a dying declaration. Its central holding — that corroboration is not a legal prerequisite — freed trial courts from a mechanical rule and returned them to the proper question: is this particular declaration true and voluntary? That is a question of credibility, to be answered on the whole record, not by a formula.
The judgment also supplied the gradation of reliability by mode of recording that practitioners still use. The magistrate's question-and-answer record, in the declarant's own words, is the gold standard precisely because it minimises the intermediary errors that can creep into a declaration reconstructed from the memory of witnesses. A declaration recorded as questions and answers shows what the declarant was actually asked and how he answered, exposing leading questions and gaps; recording it in the maker's own language preserves nuance that a paraphrase would lose; and a magistrate is a neutral officer with no stake in the prosecution. By contrast, a declaration spoken to relatives or bystanders and repeated from memory at trial is more vulnerable to honest error and to the colouring that grief or partisanship can introduce. The Court did not hold such declarations inadmissible — they remain good evidence if believed — but it placed them lower in the order of reliability. That insight informs the modern preference for magistrate-recorded declarations, even though, as the Constitution Bench would later confirm in Laxman v. State of Maharashtra (2002), a magistrate's involvement is not itself indispensable.
Read together, the two cases divide the field cleanly. Khushal Rao answers whether a dying declaration needs propping up — it does not, if true and voluntary. Laxman answers how a valid declaration may be recorded — by anyone satisfied of the declarant's fitness, with a doctor's certificate as a rule of caution rather than a precondition. Between them they form the spine of the Indian law of dying declarations.
It is important to read Khushal Rao for what it actually decided, because it is sometimes overstated. The Court did not say that every dying declaration must be accepted, nor that a declaration is to be acted upon mechanically. It said the opposite of the rigid rule it rejected: that a dying declaration is to be weighed like any other evidence, on its own merits and against the surrounding circumstances. A declaration that is inconsistent with the medical evidence, or that the declarant could not realistically have made given his injuries, or that bears signs of tutoring, may still be rejected. What the case removed was the artificial requirement of corroboration as a matter of law; what it left firmly in place was the requirement of careful judicial scrutiny as a matter of fact. The presence of three concordant declarations in this case, one of them magistrate-recorded, gave the Court the assurance of reliability that the rejected rule of corroboration had been a crude proxy for.
The successor regime under the BSA 2023
The Section 32(1) foundation of Khushal Rao migrates to Section 26 of the Bharatiya Sakshya Adhiniyam 2023, which re-enacts the provision in materially identical terms. The substantive offence moves from Section 302 of the Indian Penal Code to Section 103 of the Bharatiya Nyaya Sanhita.
The rule that a true and voluntary dying declaration needs no corroboration is a judge-made standard of appreciation, not a statutory text, and it survives the recodification untouched. After 1 July 2024, a dying declaration is admitted under Section 26 of the BSA and may sustain a conviction on its own, exactly as Khushal Rao directs. Practitioners cite the case for Section 26 declarations without qualification.
Why it endures
Khushal Rao remains the standard citation for the proposition that a dying declaration, if found true and voluntary, can be the sole basis of conviction. It is invoked in essentially every dying-declaration appeal where the defence presses for corroboration, and its gradation of reliability by mode of recording continues to guide trial courts. Decided in the Court's first decade, it set the foundation on which the whole modern law of dying declarations — including Laxman — was built.
Related on Valkya
- Laxman v. State of Maharashtra: a dying declaration needs no magistrate and no doctor's certificate
- Sharad Birdhichand Sarda v. State of Maharashtra: the panchsheel of circumstantial evidence
- State of Maharashtra v. Suresh: last seen, false explanation and the additional link
- Mukesh v. State (NCT of Delhi): the Nirbhaya appeals
Sources
- SCC OnLine — Khushal Rao v. State of Bombay, AIR 1958 SC 22 case report.
- LawBhoomi — "Landmark Supreme Court Judgements on Dying Declarations in India": https://lawbhoomi.com/landmark-supreme-court-judgements-on-dying-declarations-in-india/
- Supreme Court Observer — dying-declaration jurisprudence analysis: https://www.scobserver.in/
- LiveLaw — coverage of dying-declaration appeals: https://www.livelaw.in/
- Verdictum — Khushal Rao case digest: https://www.verdictum.in/
Related reading
Laxman v. State of Maharashtra: a dying declaration needs no magistrate and no doctor's certificate
Shivaji Sahabrao Bobade v. State of Maharashtra: "may be" guilty is not "must be" guilty
Sharad Birdhichand Sarda v. State of Maharashtra: the panchsheel of circumstantial evidence
Trace how this proposition has been treated across Indian courts — citations, bench strength, and subsequent history — in one workspace built for litigators.