The famous cases of mishandled evidence in American criminal justice share a hard lesson: physical proof is only as reliable as the people and systems handling it. Contaminated crime scenes, broken chains of custody, fabricated lab results, and forensic methods dressed up as science have produced acquittals that shocked the public and convictions that sent innocent people to prison, sometimes to death row. A 2023 analysis of 732 exoneration cases found forensic evidence errors in nearly 87 percent of them.1National Institute of Justice. The Impact of False or Misleading Forensic Evidence on Wrongful Convictions The cases below are the ones that shaped how courts, labs, and the public think about that problem.
O.J. Simpson: Blood Evidence and Chain of Custody
The 1995 murder trial of O.J. Simpson turned prosecution DNA evidence into a defense weapon. Blood found on a rear gate at the crime scene matched Simpson, and blood on socks recovered from his residence matched victim Nicole Brown Simpson.2United States Department of Justice Office of the Inspector General. USDOJ/OIG FBI Labs Report The defense argued police had planted that blood using preserved reference samples drawn from Simpson and from Nicole Brown Simpson’s body, samples stored in test tubes containing the preservative EDTA.
Two threads did the damage. First, defense attorneys pointed out that roughly 1.5 milliliters of blood from a reference vial drawn from Simpson on June 13 could not be accounted for. Second, the chain-of-custody documentation for the samples had gaps, and the FBI’s EDTA testing methodology later drew scrutiny from the Department of Justice Inspector General. Add the courtroom moment when a bloody leather glove from the scene appeared not to fit Simpson’s hand, and juror confidence in the physical evidence collapsed. Simpson was acquitted, and the case permanently changed how Americans think about police handling of crime scenes.
JonBenét Ramsey: A Contaminated Crime Scene
The 1996 murder of six-year-old JonBenét Ramsey in Boulder, Colorado, is a textbook example of contamination destroying a case before the investigation really began. When her parents reported her missing, police treated the home as a kidnapping scene rather than a potential homicide. Friends, family members, and various investigators moved freely through the house for hours, touching surfaces and rearranging objects.
Then police asked John Ramsey and a family friend to search the house themselves. John found his daughter’s body in the basement, covered by a white blanket, and carried her upstairs before investigators documented anything. That movement compromised evidence about time of death, manner of death, and any trace evidence on or near the victim. Forensic scientist Dr. Henry Lee later said that by the time investigators recognized a homicide, the scene was “totally contaminated.” The case remains unsolved nearly three decades later, and the failure to secure the scene is widely considered a primary reason.
Cameron Todd Willingham: Forensic Analysis That Was Never Valid
Cameron Todd Willingham was executed by Texas in 2004 for the arson murder of his three children. He maintained his innocence. The conviction rested on fire investigators who said they could determine the fire was intentionally set, along with a jailhouse informant claiming Willingham had confessed.
In the days before the execution, Willingham’s attorneys submitted a report from Gerald Hurst, a nationally recognized arson expert, concluding the conviction rested on erroneous forensic analysis. The governor and the Board of Pardons and Paroles declined to act. After the execution, a panel of five leading independent arson experts issued a 48-page report finding that none of the scientific analysis used to convict Willingham held up under modern fire investigation standards. In 2009, an expert hired by the Texas Forensic Science Commission concluded the original investigators should have known their analysis was wrong at the time of trial. The fire investigators had relied on indicators the arson science community has since abandoned. This case is not about contaminated samples. It is about forensic conclusions presented as science that were never scientifically valid.
Steven Avery: Allegations of Planted Evidence
Steven Avery had already served 18 years in prison for a sexual assault he did not commit before DNA evidence exonerated him in 2003. Two years later he was charged with the murder of photographer Teresa Halbach. His defense team argued that the local sheriff’s department, facing a multimillion-dollar lawsuit over the wrongful conviction, had a motive to frame him.
Avery’s blood was found inside Halbach’s vehicle. The defense pointed to a vial of Avery’s blood stored at the courthouse from his earlier case. The vial’s stopper had a small puncture hole, and the box holding it had been unsealed while in the custody of the sheriff’s department. EDTA is added to stored blood to prevent coagulation and is not naturally present in human blood, so its presence in the crime scene sample would suggest planting.3National Center for Biotechnology Information. The “Making a Murderer” Case: A Brief Description on How EDTA Is Measured in Blood The FBI tested the crime scene blood and reported no EDTA, but the defense challenged the reliability of that testing, noting the FBI had not used the protocol in over a decade and had developed it under time pressure for the trial.
Halbach’s car key raised its own questions. It was found on Avery’s bedroom floor during the seventh search of his trailer, in plain view, after six earlier searches had turned up nothing. It carried only Avery’s DNA and none of Halbach’s, despite being a key she used regularly. Avery was convicted and remains in prison.
Casey Anthony: Contested Forensic Methods
Casey Anthony was accused of killing her two-year-old daughter Caylee in 2008. Investigators had no DNA match, no confession, no murder weapon, and no definitive cause of death. The prosecution’s case leaned heavily on forensic evidence from the trunk of Anthony’s car, and that evidence split the experts.
Dr. Arpad Vass, who had developed an experimental method for analyzing chemical compounds from decomposing human remains, testified that air samples from the trunk showed chloroform at concentrations far higher than normal, consistent with a decomposing body. Michael Sigman from the National Center for Forensic Science tested the same trunk using different collection methods and concluded the results were mainly consistent with gasoline. He told jurors he could not conclusively determine that human remains had been in the trunk. The methods were still in development, and no scientific consensus existed on how to interpret the results. Caylee’s remains were not recovered until months after her disappearance, and the skeletal remains yielded almost no usable forensic information. The jury acquitted Anthony of murder. State Attorney Lawson Lamar afterward acknowledged that doubts about the forensic evidence likely added up to reasonable doubt.
Brandon Mayfield: A Fingerprint Misidentification
After the 2004 Madrid train bombings, the FBI Laboratory ran a latent fingerprint from the scene through its Automated Fingerprint Identification System. The system returned a list of 20 candidates that included Brandon Mayfield, an Oregon attorney and former Army officer. Three separate FBI examiners concluded the print matched him. The Bureau arrested Mayfield as a material witness and held him for two weeks.4United States Department of Justice Office of the Inspector General. A Review of the FBI’s Handling of the Brandon Mayfield Case
Spanish authorities had doubted the match from the beginning and eventually identified the actual source of the print as an Algerian national. The FBI released Mayfield and issued a public apology. The Department of Justice Inspector General’s investigation found the failure came from compounding errors. Examiners focused on similarities and dismissed the differences as “distortion.” Once the first examiner declared a match, the second and third were influenced by that conclusion rather than reviewing the print independently. Examiners placed excessive reliance on the AFIS output, assuming that because the computer had flagged Mayfield, he must be the source. When Spain raised concerns, the Bureau initially stood by its identification. The case led to reforms in FBI fingerprint protocols and became a widely cited example of confirmation bias in a discipline long treated as objective.
Annie Dookhan: Fabricated Lab Results at Scale
Annie Dookhan worked as a chemist at a Massachusetts state drug lab and was considered one of the lab’s most productive analysts. She processed far more samples than her colleagues and almost always confirmed the presence of illegal drugs. She was also fabricating results. Dookhan admitted to routinely failing to perform required tests, contaminating samples, forging other analysts’ initials, and sometimes identifying substances as illegal drugs without testing them at all.
She pleaded guilty to obstruction of justice, perjury, and evidence tampering in 2013 and served three years in prison. The larger reckoning came in 2017, when the Massachusetts Supreme Judicial Court ruled that more than 21,000 cases involving over 36,000 convictions tainted by Dookhan’s work had to be dismissed. It was the single largest dismissal of convictions in American history. Many affected people had already served their sentences or accepted plea deals they would never have taken had they known the testing was fraudulent. No one at the lab had checked whether Dookhan was actually running the tests she claimed. Her supervisors praised her output without questioning how one analyst could process three times the volume of her peers.
The FBI Crime Lab Investigation
In the late 1990s, the Department of Justice Inspector General completed an 18-month investigation into the FBI Laboratory, long considered the gold standard in American forensic analysis. The review covered hundreds of allegations across the Explosives Unit, the Materials Analysis Unit, and the Chemistry-Toxicology Unit.5United States Department of Justice Office of the Inspector General. USDOJ/OIG FBI Labs Report
The findings were systemic. Examiners gave scientifically flawed testimony in major cases including the 1993 World Trade Center bombing and the Oklahoma City bombing. Analysts testified beyond their expertise. Lab reports were altered, omitted key findings, or improperly supplemented. One examiner’s file review found that nearly a quarter of his cases failed to meet the lab’s own administrative and technical guidelines. The investigation was triggered largely by whistleblower Frederic Whitehurst, a supervisory special agent who had spent years raising concerns about sloppy work and biased conclusions, alleging that examiners tailored findings to support prosecution theories. The report drove reforms but also cast doubt on convictions obtained through FBI forensic testimony during the affected period.
What the Law Does When Evidence Is Mishandled
The justice system has several safeguards for evidence problems. The most important is the Supreme Court’s 1963 decision in Brady v. Maryland, which held that the prosecution’s suppression of evidence favorable to the defense violates due process when the evidence is material to guilt or punishment. The rule applies whether the suppression was intentional or accidental.6Justia Law. Brady v Maryland, 373 US 83 (1963)
When evidence is destroyed or lost rather than suppressed, the standard shifts. In Arizona v. Youngblood (1988), the Supreme Court ruled that the government’s failure to preserve potentially useful evidence does not violate due process unless the defendant can show the police acted in bad faith.7Justia Law. Arizona v Youngblood, 488 US 51 (1988) That is a high bar. Negligent or even reckless destruction of evidence, without proof of deliberate intent, often goes without constitutional remedy.
Judges have other tools. A court can give an adverse inference instruction, telling the jury it may infer that destroyed or lost evidence would have been favorable to the defense. Courts can also impose sanctions ranging from monetary penalties to excluding certain evidence, and in extreme cases they can dismiss charges entirely. These remedies depend on the defense discovering the mishandling in the first place, which does not always happen.
Before any evidence reaches a courtroom, Rule 901 of the Federal Rules of Evidence requires the party offering it to produce evidence sufficient to support a finding that the item is what they claim it is. Chain-of-custody documentation is how that requirement is met. Every person who handles a piece of evidence must be able to account for it, and the proponent must show the evidence has not been altered between collection and trial.8Legal Information Institute. Rule 901 – Authenticating or Identifying Evidence Gaps in that chain let the opposing party challenge admissibility or credibility.
How Common These Failures Are
The National Registry of Exonerations has recorded over 3,000 wrongful convictions in the United States. When researchers analyzed 732 of those cases encompassing 1,391 forensic examinations, they found errors in 891 of those examinations. The errors ranged from outright fabrication to flawed methodology to misleading testimony about what the evidence actually proved.1National Institute of Justice. The Impact of False or Misleading Forensic Evidence on Wrongful Convictions
The common thread across the cases above is institutional pressure. Crime labs answer to the same government that prosecutes cases. Fingerprint examiners know what answer investigators want. Fire investigators have long relied on training folklore rather than controlled research. Once a forensic conclusion enters a case file, the system has few mechanisms for catching the error before conviction. Reforms have followed most of these scandals, but they tend to address the last failure rather than prevent the next one.