The Fine Print on Fingerprints

In classrooms and crime scenes, Forensic Sciences Professor Heidi Eldridge relies on evidence over assumptions and encourages students to think for themselves.

As a veteran crime scene investigator, George Washington University Associate Professor of Forensic Science Heidi Eldridge has dusted for prints at bloody homicide scenes, cramped apartments and public parks. She knows some surfaces hold a clearer print than others. Dry and smooth objects like metal or glass work better than wet or textured surfaces like rain-soaked vehicles or bricks and leather.

And she knows another secret about fingerprints, perhaps forensic sciences' most iconic tool:

They aren't as foolproof as we think.

Watch any TV crime show and you'll walk away with the widely-held impression that fingerprint analysis is airtight-the "gold standard" of forensic sciences. But latent fingerprint identification in its current state is far from an exact science, explained Eldridge, who spent over a decade working crime scenes and comparing fingerprints in Oregon and Las Vegas before coming to the Department of Forensic Sciences at the Columbian College of Arts and Science (CCAS) three years ago.

Print processing suffers from shortcomings like a lack of uniform standards and the potential for human error. Indeed, it may yield a false negative error rate higher than 10%, suggested a 2011 research study by the FBI and the not-for-profit science and technology firm Noblis.

Fingerprints "are fairly accurate-but not infallible. They are not 100%," said Eldridge, who also worked as a research forensic scientist for the nonprofit RTI International. "It's not that fingerprints are unreliable. It's that we never established their reliability."

Eldridge isn't tearing down the system. Instead, she's helping rebuild it-from crime scenes to classrooms. As director of the CCAS Crime Scene Investigations Graduate Program, Eldridge teaches students the skills and techniques to succeed as forensic science professionals as well as the critical thinking that helps them make informed decisions.

A veteran crime scene investigator, Associate Professor of Forensic Science Heidi Eldridge promotes greater scientific standards for fingerprint analysis.

With a new Graduate Certificate in Latent Print Examination set to launch in fall 2026, Eldridge spoke to GW Today about training the next generation of crime scene investigators, challenging long-held assumptions and putting the "science" back in forensic science.

Q: We've always heard that fingerprints are the foundation of crime scene investigations. Is that wrong? Are they not reliable?

A: They are not as reliable as we thought. They are still very reliable. But they are not the gold standard. DNA is now referred to as the gold standard. In fact, when DNA was in its infancy in the '90s, it was called "DNA fingerprinting" to borrow that legitimacy from the fingerprint field.

Fingerprints are still the workhorse of forensic identification because they're faster, cheaper and easier to collect than DNA. But we never established the reliability of fingerprints. Now we're going back and finding that, in general, fingerprint examiners are pretty darn good at what we do. But we've been going into court saying, "This method has a zero-error rate." And it definitely does not.

Q: Do we know what the error rate is?

A: We have a ballpark estimate from self-selected examiners who volunteered to participate in studies. From those, we've seen false positive rates hovering around 0.1%. The false negative rate is around 8-to-10%.

Q: That sounds like a very high number. Are we saying fingerprint identifications are wrong 10 percent of the time?

A: No, not exactly. We're saying the false negative rate is 10%. That doesn't mean we are identifying people incorrectly. It means we are not identifying some people at all.

So let's say I find a latent print at a crime scene. I compare it to a suspect and I conclude the suspect did not leave this print-and I'm wrong. That's a false negative.

In the criminal justice system, we consider that to be a lesser kind of mistake. We say it is worse to falsely accuse somebody than to let a guilty person go free. But it's still a problem. If guilty people go free, they may commit more crimes.

Q: How did the idea of fingerprint infallibility start? Was there science to support it? Or is it a crime show creation?

A: Well, TV hasn't helped. At the start of every semester, I ask students what got them interested in forensic science. Almost everyone names a TV show. I hear a lot of "I have memories of watching 'Bones' with my mom when I was a kid."

But this pre-dates the "CSI" era. There was little foundational research in latent prints because it grew up as a police science. Back in the day, police had the bright idea to look at these fingerprints and see if they were useful-and they seemed to work. But they didn't go through the rigorous scientific process of experimentation and building a foundation to show that it was actually a reliable method. And now, we say, "We have been accepting prints for 100 years so, of course, they are reliable." The law relies on precedent. That may be a decent legal argument, but it's a lousy scientific one.

Q: Is the unreliability issue a problem with technology or people?

A: It's human error. There isn't really a lot of technology where these decisions are involved.

If you go into drug chemistry, you're using gas chromatograph mass spectrometers. If you go into DNA, you've got sequencers and all sorts of technology.

If you go into latent fingerprints, your instrument is the human brain.

We look at prints, we see what they have in common, and we make a judgment call. And that process is fraught with places where we can make mistakes.

Q: There is no advanced technology for examining fingerprints?

A: The main technology we have is AFIS, the Automated Fingerprint Identification System.

[NOTE: AFIS is a computer system that uses specialized software and algorithms to compare fingerprints against databases to identify individuals.]

But AFIS is a search tool. I give it a print and it gives me a ranked list of candidates who might be close matches. A person still has to compare each match, usually using imaging software like Photoshop. But there is no software and no computer that says, "Ding! This is the guy we're looking for."

Q: Can we improve on it?

A: Absolutely. We cannot 100% fix it. There is always going to be a human examiner with human judgment and the opportunity for human error. But we can identify our weaknesses through research. And we can use that data to inform the creation of stringent standards for each step of this process. Part of those standards would be documentation requirements. With good documentation, it will be easier to go back and review our mistakes to understand what went wrong and how to correct the problem.

Graduate students in Eldridge's Crime Scene Investigations I class practice a crime scene mapping technique to measure the location of found evidence at an outdoor scene.

Q: What happens in our classrooms? Are we giving forensic sciences students the kind of scientific foundation they need?

A: Yes. And they are benefiting from the real world experience of people like me and [Department Chair] Daniele Podiniwho have worked in the field.

Critical thinking is also an important part of our program. So much of what was done for the last 100 years was taken on faith. People weren't criticizing or questioning the status quo. We're teaching students to advocate for how things should be done and to challenge their agencies to be better. We want them to ask: Why aren't we documenting? Why aren't we following standards? Why aren't we working in a way that's more defensible in court?

Q: What's the lesson you want students to take with them when they leave here and become forensic science professionals?

A: I want them to have all the foundational skills necessary to start off their jobs not as average trainees but as trainees who already know the basics and have a leg up.

And I want them to have a realistic view of what they're going to be doing. This is a hard job. TV glamorizes it. But you're working crazy hours, you're getting woken up in the middle of the night, you're in physically uncomfortable conditions. You're seeing things that you can't unsee.

It's hard to see people on their worst days, to see bodies lying in pools of blood, to go to autopsies and smell a dead body that's been fished out of a river. It's not what people necessarily thought about when they pictured themselves fighting crime. It can be a shock. One of the best things we can do for students is to prepare them for that reality.