Dr David Kelly, the weapons expert at the center of the Iraq dossier row, did not take his own life but was murdered to silence him, and the official suicide finding is a cover-up

The case people make

The doubt in this case deserves to be stated at its strongest, because unlike many conspiracy claims it did not come only from the fringe. It came, in significant part, from doctors.

A group of medical professionals, some of them specialists, went on the record to argue that the stated cause of death did not add up. The severed vessel in Kelly's wrist, the ulnar artery, is small, and they contended that a cut to it would be unlikely to produce a fatal loss of blood: the artery would tend to retract and clot rather than bleed out. They noted, too, that the level of co-proxamol measured in his body was below what is usually considered lethal. If neither the wound nor the drugs would ordinarily kill, they asked, how could this be a straightforward suicide?

Around that medical core gathered a set of genuine grievances about process. No full coroner's inquest was ever held; a government-appointed inquiry took its place, and some post-mortem records were placed under a lengthy restriction. To critics this looked like officialdom keeping control of an account it did not want tested. Add the context, a weapons expert dead at the very height of a scandal over whether the public had been misled into war, and the impulse to ask hard questions was not paranoia but a reasonable response to an extraordinary situation.

When qualified doctors say the medical cause does not hold, and the state answers by replacing an inquest with an inquiry, the demand for a proper hearing is not a conspiracy theory. It is a request for the ordinary machinery of scrutiny to be allowed to run.

That is the honest form of the case: not that a murder has been shown, but that the death occurred in charged circumstances, that credible people found the official explanation medically unconvincing, and that the way the matter was handled left legitimate questions without the full public airing they might otherwise have had.

The medical dispute, examined

The medical objection is the load-bearing part of the case, so it deserves to be met directly rather than waved away. It is also where the argument turns out to be more evenly balanced than the headline version suggests.

The claim that a cut ulnar artery “could not” be fatal treats the wound in isolation, but the official finding never rested on the wound alone. It rested on a combination: the incised wounds, an overdose of co-proxamol, and significant, previously undiagnosed coronary artery disease. The pathologist, and later two independent experts commissioned during the Attorney General's review, took the view that a diseased heart and a drug overdose together made Kelly far more vulnerable to the effects of blood loss than a healthy person would have been, so that a bleed which might not kill on its own could do so in his particular condition. The same logic answers the point about the drug level: co-proxamol was treated as a contributor, not the sole agent.

This does not mean the doctors who objected were foolish or acting in bad faith. Their reading is a defensible one, and the disagreement between competent experts is real. But a genuine difference of medical opinionabout how three factors combined to cause a death is a very different thing from evidence that the death was a homicide. The dispute, at its height, establishes that the cause of death is arguable. It does not identify a killer, a weapon in someone else's hand, or any positive sign that a third party was present.

It is worth being plain about the gap the theory has to cross. The medical cause is contested and therefore he was murdered are not the same statement, and the second does not follow from the first. Doubt about a diagnosis is an argument for more scrutiny of the diagnosis. It is not, by itself, an argument that anyone killed him.

Where the murder claim runs out

Set the medical argument to one side and ask what positive evidence exists that Kelly was murdered. This is where the case thins to almost nothing.

After more than twenty years, no one has produced a killer, a method, a motive traced to a real actor, or a single witness to foul play. The strongest version of the theory runs on the absence of a satisfying explanation rather than on the presence of an incriminating one. That is a structural weakness: a death can be genuinely hard to explain, and can have been mishandled by the authorities, and still not be a homicide. Hard-to-explain is a property of many suicides and accidents, not a fingerprint of murder.

The theory's pillars, examined one by one, do not carry the weight placed on them. Motive is not act: that Kelly was inconvenient to powerful people explains why the suspicion is attractive, not that any of them did anything, and, notably, he had not recanted or threatened to expose a further secret; he had given his account and expected more of the same ordeal. Process failures are not proof: the missing inquest and the restricted records are fair criticisms of how the state behaved, but the material was later opened and re-examined. Scene impressions are not measurements: recollections of how much blood was visible are not the forensic record, which the reviewers had in full.

And the claim has been tested more than once. In 2011 the Attorney General, Dominic Grieve, reviewed the calls for an inquest, obtained fresh opinions from an independent forensic pathologist and a toxicologist, and concluded that the evidence Kelly took his own life was overwhelmingly strong and that nothing he had seen supported an allegation of murder or cover-up. He declined to seek a new inquest, and a subsequent legal challenge to that decision failed. Two separate examinations, years apart, landed in the same place.

Why the doubt endures

More interesting than whether the theory is true is why it has been so durable, and so much more respectable than most, and the answer lies in the circumstances rather than in any hidden proof.

It rests on a foundation of real facts. The doubts were voiced by credentialed doctors, the process really was unusual, and the political context really was as charged as it looks. When the starting materials are all genuine, the final leap, from this is troubling to this was murder, feels shorter than it is, because so much of the ground beneath it is solid.

It is bound to the Iraq war. The wider case for the war was built on weapons claims that proved false, a story told in its own file on this site, and a government caught out on its central justification forfeits the benefit of the doubt on everything nearby. To a public that had watched the dossier collapse, official reassurance about the death of the dossier's critic was always going to be a hard sell, regardless of its merits.

And it answers a human difficulty. Those who knew Kelly describe a capable, dignified man, and it is genuinely hard to absorb that such a person, humiliated and cornered, might have been driven to despair. An external villain can feel more proportionate to the loss, and, in a strange way, easier to bear than the alternative. The endurance of the theory is partly a measure of how much this death mattered, and how reluctant people are to accept the ordinary, sadder reading of it.