Evidence in forensic science is any physical, biological, or digital material recovered from a crime scene that can help establish what happened and who was involved. The entire field rests on a simple idea known as Locard’s Exchange Principle: every contact leaves a trace. When a person enters a space, they bring something with them and leave something behind, whether that’s a fingerprint, a clothing fiber, a hair, or data on a phone. Forensic science is the process of finding, collecting, analyzing, and presenting those traces in a way that holds up in court.
Physical vs. Testimonial Evidence
Forensic evidence falls into two broad categories. Testimonial evidence is the spoken word, meaning statements from victims or witnesses. Physical evidence, sometimes called real evidence, is anything tangible recovered from a scene. The National Institute of Justice lists examples including blood, semen, saliva, fibers, paint chips, glass, soil, fingerprints, hair, narcotics, shoe prints, tire tracks, tool marks, and fracture patterns in materials like glass or adhesive tape.
Within physical evidence, forensic scientists further distinguish between biological evidence (anything from a living organism, like blood or skin cells), trace evidence (microscopic materials transferred during contact), impression evidence (patterns left by objects pressing into surfaces), and digital evidence (information stored on or transmitted by electronic devices). Each type requires different collection methods, different lab techniques, and sometimes different legal standards.
Biological Evidence and DNA
Biological evidence is often the most powerful link between a suspect and a crime scene. Any sample containing cells with a nucleus, including liquid blood, dried bloodstains, saliva, semen, and skin cells, can potentially yield a DNA profile. The lab process involves four main steps: extracting DNA from the sample, measuring how much DNA is present, amplifying it to create enough material for testing, and then detecting the resulting genetic markers.
When samples are degraded by heat, moisture, or bacteria, analysts need more cellular material to build a usable profile. In sexual assault cases, Y-chromosome analysis allows investigators to isolate the male genetic contribution from a mixed sample. Mitochondrial DNA, which is inherited only from the mother, is useful when nuclear DNA is too damaged. It can also link evidence to a maternal family line, making it valuable for identifying remains.
Trace Evidence
Trace evidence refers to the tiny, often invisible materials transferred when two surfaces meet. Traditional categories include glass fragments, fibers from clothing or carpet, tape, paints and dyes, gunshot residue, ignitable liquids, explosive residues, minerals and soils, and even pollen. A single carpet fiber on a suspect’s shoe or a flake of automotive paint at a hit-and-run scene can connect a person or vehicle to a specific location.
Because these materials are so small, collection requires meticulous technique. Analysts use tools like tape lifts, vacuum filters, and tweezers under magnification. In the lab, techniques such as infrared spectroscopy, Raman spectroscopy, and elemental analysis help determine whether a recovered fiber or paint chip matches a known source. The goal is rarely to prove an exact one-to-one match. Instead, trace evidence narrows the possibilities, placing a suspect in a category of people or objects consistent with the scene.
Impression Evidence
Fingerprints, shoe prints, tire tracks, and tool marks are all forms of impression evidence. They work on the same principle: a harder surface presses into or slides against a softer one, leaving a pattern. Firearms evidence falls into this category too. When a gun is fired, the barrel leaves unique scratches (called striations) on the bullet, and the firing mechanism stamps distinctive marks onto the cartridge casing. Examiners compare those marks against test-fired rounds from a recovered weapon to determine whether it was the gun used in a shooting.
Newer technology is making these comparisons faster and more standardized. Three-dimensional imaging sensors can capture the surface of a bullet or casing in roughly two minutes, and the National Institute of Standards and Technology is building an open-access ballistics database that researchers and software developers can use to improve automated matching algorithms.
Digital Evidence
Digital evidence has become central to modern investigations. Phones, computers, cameras, and cloud accounts all contain data that can place a person at a specific location at a specific time or reveal communications relevant to a case. One of the most useful forms of digital evidence is metadata: the hidden information embedded in every file.
A digital photograph, for example, can contain the device model that took it, the exact date and time, GPS coordinates, and a record of any edits made afterward. Documents carry similar data, including the author’s name, creation and modification dates, the software used, and file size. Forensic analysts can recover file names, access histories, records of program executions, and even data the user thought they deleted. Metadata also helps authenticate evidence by revealing whether a file has been altered or fabricated.
Chain of Custody
None of this evidence matters in court if it cannot be proven authentic. That is the purpose of the chain of custody, which the National Library of Medicine describes as the most critical process in evidence documentation. It is a sequential record that tracks every person who handled a piece of evidence, from the moment of collection through analysis and into the courtroom.
Each evidence container gets a unique identification code along with the location, date, and time of collection, the collector’s name and signature, and a witness signature. Every time evidence changes hands, the new custodian signs, dates, and notes the time on a chain of custody form. That form also records the method of delivery, the type of analysis requested, and storage conditions. Evidence must be sealed in tamper-resistant bags or with tamper-evident tape, and a separate chain of custody form accompanies each evidence bag. If there is any gap in this paper trail, a defense attorney can argue the evidence was compromised, potentially making it inadmissible.
How Courts Decide What Evidence Is Reliable
Before forensic evidence reaches a jury, a judge must decide whether the science behind it is sound. In most federal courts and many state courts, this is governed by the Daubert standard, which replaced the older Frye standard. Under Frye, the only question was whether a technique had “general acceptance” in its scientific field. Daubert, established by the Supreme Court in 1993, is more demanding. It asks five questions about any expert methodology: Can it be tested, and has it been? Has it been peer-reviewed and published? What is its known or potential error rate? Do standards exist for how the technique is applied? And is it generally accepted in the relevant scientific community?
In 1999, the Supreme Court extended these criteria beyond hard science to include testimony from engineers and other technical experts. This means forensic disciplines like tool mark analysis and handwriting comparison face the same scrutiny as DNA profiling.
Known Limitations and Error Rates
Not all forensic evidence carries equal scientific weight. A landmark 2009 report from the National Research Council concluded that, with the exception of nuclear DNA analysis, no forensic method had been rigorously shown to consistently link evidence to a specific individual with a high degree of certainty. A follow-up review by the President’s Council of Advisors on Science and Technology in 2016 reached similar conclusions.
The challenge is measuring how often forensic methods produce wrong answers. Early studies of pattern-matching disciplines like firearms analysis used designs that made errors artificially unlikely, and those flaws are now widely recognized. Only five studies to date have used more realistic experimental designs, and they report false positive error rates of 1 to 2%. For firearms and tool mark examination specifically, independent validation remains limited: researchers have generally failed to provide enough methodological detail for others to replicate their work, and most published studies come from practicing examiners within the field rather than independent scientists.
This does not mean these disciplines are useless. It means the scientific foundation beneath some forensic methods is still being built, and courts are increasingly aware of the difference between a DNA match (which can narrow the source to one in billions) and a subjective pattern comparison where error rates are harder to pin down. Understanding that distinction is essential for anyone trying to evaluate what forensic evidence actually proves.