Fingerprint evidence has convicted people for over a century and remains one of forensic science's most trusted tools. But "trusted" and "infallible" are different things, and the modern, honest answer to how reliable is it? is more nuanced than television suggests.
What the reviews found
Two influential reports pushed the field toward candour. The US National Academy of Sciences report (2009) criticised many pattern-comparison disciplines for lacking measured error rates and rigorous validation. The PCAST report (2016) went further, calling for empirical "black-box" studies to establish how often examiners actually get it wrong.
What the error-rate studies show
Large black-box studies, in which examiners judge prints of known origin, have measured performance directly. The headline findings:
- false positives (wrongly identifying) are rare but not zero — on the order of a fraction of a percent,
- false negatives (missing a true match) are somewhat more common,
- and examiners are generally well-calibrated when they call something inconclusive.
So the method is highly accurate — but it has a real, measurable error rate, driven mostly by poor-quality marks and human factors.
How it should be presented
The reforms following Mayfield and these reviews point to a clear standard for testimony:
- avoid claims of 100% certainty or identification "to the exclusion of all others in the world",
- acknowledge the discipline's error rate,
- ensure conclusions are documented, verified and, ideally, bias-mitigated.
The bottom line
Fingerprint evidence deserves its strong reputation — but as expert opinion with a known, small error rate, not as mechanical proof. Courts (and juries) are best served when its power and its limits are stated plainly.