DNA evidence is extraordinarily accurate when a clean, single-source sample is properly collected and analyzed. A standard DNA profile using 20 genetic markers has a random match probability of roughly 1 in 10 trillion or better, making it virtually impossible for two unrelated people to share the same profile. But “accurate” in a courtroom means more than just the science of matching. The chain from crime scene to conviction involves sample collection, laboratory processing, software interpretation, and human judgment, and errors can enter at every stage.
What Makes DNA Matching So Precise
Forensic DNA profiling works by examining short, repeating sequences at specific locations on your chromosomes. These locations are called loci. The FBI’s national database, CODIS, originally required profiles at 13 loci. In January 2017, that requirement expanded to 20 core loci, making modern profiles far more discriminating than earlier ones.
At 13 loci, the chance that a random person’s DNA would match someone else’s profile is about 1 in 10 trillion. With 20 loci, that probability shrinks even further. For context, there are fewer than 8 billion people on Earth. So when analysts get a full, clean profile from a single person, the identification power of DNA is essentially unmatched by any other forensic method.
Where Errors Actually Happen
The math behind DNA matching is rock solid. The problems arise in the human and procedural steps surrounding it. A review of the Netherlands Forensic Institute found that general DNA laboratory errors (contamination, sample mix-ups, technical mistakes) occurred in less than 1% of all analyses, and most involved human error rather than scientific failure. That sounds reassuring, but forensic labs process enormous volumes of cases, so even a sub-1% error rate can affect real people.
The types of errors that matter most include mislabeled samples, contamination during collection or handling, and misinterpretation of results. A swab taken carelessly at a crime scene can pick up DNA from a bystander or a first responder. A lab technician’s own DNA can contaminate a sample. These aren’t failures of the science itself, but they can produce results that point to the wrong person.
Published research on real-world error rates in forensic DNA analysis remains surprisingly thin. A survey of forensic science literature noted a persistent lack of published data on the rate of erroneous conclusions in DNA analyses. Most existing studies use controlled conditions that don’t closely resemble actual casework, limiting how much we can generalize from them.
Mixed Samples Create Real Problems
DNA evidence becomes significantly less reliable when a sample contains genetic material from more than one person. This is common: a doorknob, a weapon, or clothing frequently carries DNA from multiple people. Two-person mixtures are generally manageable for trained analysts, provided the DNA concentrations are above a minimum threshold. Three-person mixtures, however, are a different story. The National Institute of Justice found that three-person mixtures “are generally beyond the scope of protocol limits for most participating examiners,” and the majority of labs in their study had difficulty interpreting them.
To handle complex mixtures, forensic labs increasingly rely on probabilistic genotyping software like STRmix. These programs use statistical modeling to estimate the likelihood that a specific person contributed to a mixed sample. Validation studies show the software generally performs well, but it isn’t perfect. In rare cases, STRmix has excluded a true contributor from a mixture because of random fluctuations during the amplification process. The software produces a likelihood ratio rather than a yes-or-no answer, and interpreting that ratio still requires human judgment.
Your DNA Can Be Somewhere You’ve Never Been
One of the most important limitations of DNA evidence has nothing to do with lab accuracy. Secondary DNA transfer means your genetic material can end up at a location you never visited. If you shake someone’s hand and that person later touches an object at a crime scene, your DNA could be recovered from that object. A 2023 literature review confirmed that secondary transfer “is possible and may result in a single full profile,” meaning investigators could find what looks like a complete DNA match from someone who was never there.
Several factors influence how much DNA transfers indirectly. People vary widely in how much DNA they shed; so-called “good shedders” leave behind more genetic material, which is then more likely to survive a second transfer. The type and duration of contact matters too, with longer or more intimate contact depositing more DNA. Moisture, surface texture, and the original source of DNA (saliva transfers more effectively than dry skin cells, for example) all play a role. In some experiments, the good shedder’s profile made up 80 to 100% of the DNA recovered after secondary transfer, potentially overwhelming any trace from the person who was actually present.
This means DNA evidence can reliably tell you whose genetic material is on an object, but it cannot tell you when or how it got there. That distinction is critical in court.
Environmental Degradation
DNA left at a crime scene doesn’t stay intact forever. Sunlight, heat, moisture, and microbial activity all break down genetic material over time. Ultraviolet radiation from sunlight distorts DNA’s structure by causing adjacent bases to fuse together. Moisture accelerates a process called hydrolysis, which severs the chemical bonds holding DNA’s building blocks together. Heat speeds up both of these reactions.
Degraded DNA doesn’t necessarily produce a wrong result, but it can produce an incomplete one. When DNA breaks into smaller fragments, some of the 20 loci may fail to amplify, yielding a partial profile. A partial profile is less discriminating. Instead of a 1-in-10-trillion match probability, a profile with only 8 or 10 usable loci might have match odds in the millions or billions. Still powerful, but the door opens wider for coincidental matches, especially when searching large databases.
How Courts Evaluate DNA Evidence
In federal courts and most state courts, DNA evidence must pass the Daubert standard before a jury hears it. Under this framework, the trial judge acts as a gatekeeper, evaluating whether the scientific methods behind the evidence are testable, have known error rates, have been peer-reviewed, and are generally accepted in the scientific community. DNA profiling clears these hurdles comfortably for single-source samples. For complex mixtures or degraded samples interpreted with probabilistic software, the admissibility question gets more contested, and defense attorneys increasingly challenge the reliability of these analyses.
DNA evidence has also been one of the most powerful tools for proving innocence. In 2024 alone, 147 people were exonerated in the United States after losing an average of 13.5 years to wrongful imprisonment. Across those cases, 29% involved false or misleading forensic evidence as a contributing factor. DNA testing, applied after the fact, has been the single most effective technology for uncovering these mistakes, demonstrating both the power of accurate DNA analysis and the consequences when forensic evidence of any kind is misapplied.
What “Accurate” Really Means
DNA evidence is best understood as two separate questions. The first is whether two DNA profiles match, which modern technology answers with near-perfect precision for clean, single-source samples. The second is what that match means in context: whether the DNA arrived through direct contact, secondary transfer, or contamination, and whether the sample was degraded, mixed, or mishandled along the way. The science behind the first question is among the most reliable in all of forensics. The second question is where uncertainty lives, and it depends on the specific circumstances of each case.