A murder conviction is supposed to mark the end of doubt. Yet some of the most instructive case files in criminal justice are the ones that later unravel. These examples of overturned murder convictions matter because they show, in practical terms, how an investigation can go wrong, how a jury can be misled, and how weak reasoning can harden into a verdict.

For anyone interested in miscarriages of justice, the pattern is familiar. A suspect is identified early, police theory narrows, contradictory material is discounted, and forensic evidence is presented with more certainty than it deserves. Years later, the conviction falls apart under fresh science, withheld material, or a more disciplined review of the evidence.

Why examples of overturned murder convictions matter

An overturned conviction is not just a dramatic legal event. It is a diagnostic tool. It shows where the system failed - at the crime scene, in interview rooms, in forensic interpretation, in disclosure, or in the courtroom itself.

That is why these cases deserve careful reading rather than quick reactions. Some involve deliberate misconduct. Others reflect ordinary investigative error made dangerous by momentum, pressure and overconfidence. The distinction matters, but so does the result: the wrong person convicted, while the truth is delayed or lost altogether.

1. Andrew Mallard

For an Australian audience, Andrew Mallard remains one of the clearest warnings about tunnel vision. Mallard was convicted over the 1994 murder of Pamela Lawrence in Western Australia. The prosecution case relied heavily on what was framed as incriminating knowledge in his police interview, despite the fact he was vulnerable, suggestible and not represented during questioning.

The conviction was eventually quashed after serious concerns emerged about undisclosed material and the way the case had been built. The High Court found the prosecution had failed in its disclosure obligations. In plain terms, the defence had not been given material that could have assisted Mallard.

The case shows how dangerous it is when investigators become fixed on one suspect and treat ambiguous statements as confirmation. It also underlines a basic point: disclosure is not procedural housekeeping. It goes to whether a trial can be fair at all.

2. Lindy and Michael Chamberlain

The Chamberlain case is one of the most well-known Australian examples of a wrongful conviction connected to a death initially treated as murder. Lindy Chamberlain was convicted over the death of baby Azaria in 1982, largely because the prosecution persuaded the court that supposed forensic signs pointed to homicide rather than a dingo attack.

What followed was a long and painful correction. Key forensic claims were later discredited. Material thought to be blood staining was found to have innocent explanations, and later discoveries supported the possibility that a dingo had taken the child, as the Chamberlains had maintained from the beginning.

This case is still studied because it shows how forensic evidence can appear powerful while resting on poor science or overstatement. It also shows the effect of public prejudice. Once a narrative takes hold, unusual behaviour, religious difference, or perceived lack of emotion can be read as guilt when they are nothing of the sort.

3. Henry Keogh

Henry Keogh was convicted in South Australia over the death of his fiancée, Anna-Jane Cheney, whose body was found in a bath in 1994. The prosecution relied heavily on expert pathology evidence that suggested assault and drowning rather than accidental death.

Years later, that forensic evidence came under sustained challenge. The reliability of the pathology opinions was central. Keogh’s conviction was ultimately overturned, and he was later acquitted after a retrial was abandoned.

The lesson here is not that expert evidence is unhelpful. Far from it. The lesson is that expert evidence must stay within the limits of what the science can actually support. Once an expert opinion hardens from possibility into certainty, juries can be given a false sense that the case is stronger than it really is.

4. David Eastman

David Eastman was convicted over the 1989 murder of Australian Federal Police Assistant Commissioner Colin Winchester. It was a major case, politically charged and intensely scrutinised. Eastman spent many years in prison before his conviction was quashed.

The problems in the case were broad and serious. They included concerns about witness reliability, the conduct of the investigation, and the handling of forensic and psychiatric evidence. After a long inquiry and further proceedings, Eastman was acquitted at retrial.

This is one of those cases where a single neat explanation does not do justice to the record. Sometimes a wrongful conviction is not built on one spectacular failure. It is built on layers of weakness that, taken together, should never have supported a verdict beyond reasonable doubt.

5. Barry George

In the United Kingdom, Barry George was convicted of the murder of television presenter Jill Dando. The case drew enormous publicity, and the conviction initially appeared to settle a national mystery. But the evidence was always more fragile than the headline version suggested.

A tiny particle said to be firearm discharge residue played a major role at trial. Later review cast doubt on the significance and reliability of that evidence. George’s conviction was quashed, and at retrial he was acquitted.

What this case demonstrates is the danger of attaching too much weight to a narrow forensic point, particularly in a high-profile matter where the pressure to solve the case is intense. A small piece of science can end up carrying a burden it was never fit to bear.

6. Michael Morton

In the United States, Michael Morton was wrongfully convicted over the 1986 murder of his wife in Texas. He spent nearly 25 years in prison before DNA testing identified another man as the likely offender.

Morton’s case is often cited because it involved the suppression of evidence that should have been disclosed to the defence. Material indicating that a child witness had given information inconsistent with the prosecution theory was not properly handed over. Neither was other exculpatory material.

This is one of the clearest modern examples of how disclosure failures do not simply weaken confidence in a conviction - they can produce one. If the defence is denied access to facts that point away from guilt, the trial itself becomes structurally distorted.

7. Juan Rivera

Juan Rivera was convicted in Illinois over the rape and murder of an 11-year-old girl. His case involved repeated trials, multiple reversals, and a false confession that sat at the centre of the prosecution narrative despite strong reasons to doubt it.

DNA evidence excluded Rivera, yet the case persisted for years. That fact alone tells you how powerful confession evidence can be, even when it is inconsistent, contaminated by police knowledge, or contradicted by objective science.

Investigators know that a confession is persuasive. Juries know it too. But not all confessions are reliable. Fatigue, pressure, vulnerability, suggestive interviewing and a suspect’s psychological state can all produce statements that look damning on paper and collapse under proper scrutiny.

8. The Norfolk Four

The Norfolk Four case arose from the rape and murder of Michelle Moore-Bosko in Virginia. Four men were convicted, largely on the basis of confessions obtained through aggressive interrogation. The confessions were inconsistent with one another and with parts of the physical evidence.

Over time, the case became a textbook example of how multiple false confessions can emerge in the same investigation. That may sound improbable to those outside criminal casework, but once investigators feed details, reject denials and reward agreement, a distorted account can be built repeatedly.

The eventual pardons and legal reversals did not erase the years lost. What the case leaves behind is a blunt warning: when police interviews become exercises in confirmation rather than testing, the risk to truth is immediate.

What these overturned murder convictions have in common

Across these examples of overturned murder convictions, the recurring failures are strikingly consistent. The first is tunnel vision. Once investigators settle on a suspect, all later material is too often sorted into two piles: evidence that fits and evidence that gets explained away.

The second is overstated forensic science. Forensics can be decisive, but it can also be uncertain, contaminated, poorly interpreted or presented to a jury with more confidence than the discipline allows. That gap between science and courtroom language has done real damage.

The third is interview risk. False confessions are not a fringe issue. They occur in cases involving youth, cognitive vulnerability, exhaustion, intimidation and poor interviewing practice. A recorded interview helps, but only if the full process is fair and transparent.

The fourth is disclosure. A criminal trial depends on both sides having access to relevant material. When important information is withheld, whether deliberately or through neglect, the entire fact-finding process is compromised.

For those who follow justice issues seriously, including readers of Graeme Crowley Investigates, these cases are not reasons to lose faith in the system altogether. They are reasons to examine it properly. Good investigations test assumptions, preserve doubt where doubt is warranted, and remain open to being wrong. That is not weakness. In murder cases, it is discipline.

If there is one useful habit to take from these files, it is this: pay close attention to how the case was built, not just how it was sold. That is usually where the truth starts to reappear.