Indian prime minister Lal Bahadur Shastri's sudden death in Tashkent in 1966, officially a heart attack, was suspected by many to be a poisoning, a suspicion kept alive by an absent autopsy and by the state's continuing refusal to open its file

The autopsy that was never done

Everything about this case turns on a single omission. When a serving prime minister dies suddenly on foreign soil, the expectation is that the state will examine the body exhaustively. That did not happen. No post-mortem was performed in Tashkent, and none was performed after the body reached India. The verdict of a natural death rests entirely on the clinical judgment of the doctors present, not on any forensic or toxicological record.

Into that vacuum came the family's observations. Shastri's son Anil Shastrihas said the body had turned blue and bore white marks, and that the family was told such signs can accompany poisoning. Shastri's widow, Lalita Shastri, is reported to have asked why the body was discolored and about certain marks on it. The journalist Kuldip Nayar, who was part of the delegation and later Shastri's biographer, recorded those doubts as they were voiced at the time.

None of this is proof. Cyanosis and post-mortem changes occur in cardiac deaths and can be produced or exaggerated by the handling, transport, and embalming of a body. But the point the theory presses is fair on its own terms: with an autopsy, these impressions could have been tested and either confirmed or laid to rest. Without one, they can only hang in the air.

A post-mortem in January 1966 would very likely have answered the question for good. Its absence is why the question is still asked.

The file the state will not open

The second engine of suspicion is official secrecy. Over the years, citizens have used the Right to Information Act to ask what the government knows about the death. The Prime Minister's Office has said that it holds a single documentrelating to Shastri's death and has declined to release it, invoking the exemption for information whose disclosure could prejudicially affect relations with a foreign state.

That refusal has been challenged. In 2018 the Central Information Commission, through Information Commissioner M. Sridhar Acharyulu, directed the PMO and the ministries of external affairs and home to state what records they hold, and held that the government has a duty to explainhow Shastri died and that the public's right to know the truth cannot be brushed aside. Alongside this sits the puzzle of the 1977 inquiry associated with Raj Narain, whose records are reported to be untraceable or unpublished.

To believers, a government that admits holding exactly one file and still will not open it is all but confessing. That reading is understandable, and the grievance about opacity is legitimate. But it is worth stating plainly what the secrecy does and does not show. It shows that the state is withholding a diplomatic record touching a foreign power. It does not show what is in that record, and withholding is consistent with embarrassment or protocol as much as with the cover-up of a killing. A closed file is a reason to keep asking; it is not, by itself, an answer.

What the evidence does not show

Set against the doubts is the plainer possibility that the official cause is simply correct. Shastri had a history of heart trouble and had just carried the strain of leading a nation through war and then through tense, round-the-clock negotiations, in an unfamiliar city and a cold January. A fatal heart attack in those conditions needs no assassin. On the known medical facts it is the most probable explanation, which is why it remains the official one.

The affirmative case for poisoning, examined closely, thins out. The discoloration and marks are unrecorded impressions, never documented by autopsy or toxicology. The stories of witnesses later dying in accidents are anecdotal and have never been tied to the death by any inquiry. The sealed file is real but its contents are unknown, so it cannot support a specific charge. What the theory has is a powerful set of absences: no post-mortem, no open record, no published inquiry. Absences are fertile ground for suspicion, but they are not the same as evidence of a crime.

That is the honest shape of the case. There is no proof Shastri was poisoned. There is also no autopsy that could have ruled it out, and no fully open file that could put the matter to rest. The theory is unproven, and it is unproven in the specific, frustrating way that a death without a forensic record always will be.

The case for poisoning is built almost entirely from what is missing. Missing evidence keeps a question open; it does not settle it.

Where the evidence lands

Keep the layers apart. The event is documented: Shastri died in Tashkent in the early hours of 11 January 1966, hours after signing the declaration that ended the 1965 war, and the cause on the record is a heart attack. The poisoning theory is unproven: no post-mortem was ever done, no forensic evidence of poisoning exists, and the family's observations, however sincere, were never converted into clinical findings.

What sustains the doubt is not positive proof of murder but two real deficits: the absent autopsy and the state's refusal to open its single file, a refusal the Central Information Commission has itself criticized. Those deficits are legitimate grievances. They explain why a natural death has never felt settled to Shastri's family or to much of the public, and why the question keeps returning at each anniversary and each RTI ruling.

The responsible posture is to report both truths without collapsing one into the other. Shastri may well have died exactly as the record says. He may not have. What can be said with confidence is that the way his death was handled, no examination of the body, no durable public inquiry, no open file, guaranteed that the question could never be closed. This file rates the poisoning claim unproven, and treats the enduring suspicion as the predictable consequence of secrecy and a missing autopsy, not as a demonstrated fact.