Ballistic fingerprinting works by matching the microscopic marks a firearm stamps onto bullets and cartridge casings when it fires. Every gun’s internal surfaces carry tiny imperfections from manufacturing, wear, and corrosion, and those imperfections transfer onto the softer metal of ammunition components each time the weapon is used. A forensic examiner captures those marks from crime scene evidence, test-fires a suspect firearm to produce a known reference, and compares the two under a specialized microscope to decide whether the same gun produced both.
How a Firearm Marks a Bullet and Casing
Several parts of a gun touch the ammunition with enough force to leave impressions. The firing pin strikes the primer at the base of the cartridge, leaving an indentation shaped by that pin’s geometry. The explosion in the chamber drives the casing backward against the breech face, pressing the breech’s surface texture into the casing. As the action cycles, the extractor pulls the casing out of the chamber and the ejector kicks it clear, each leaving its own scratches. The bullet, meanwhile, travels down the barrel, where spiral grooves called rifling carve a pattern of raised and recessed impressions into the projectile.
Those marks are what an examiner reads. A casing carries firing pin, breech face, extractor, and ejector marks. A bullet carries rifling impressions along its bearing surface. Damage to either surface after the shot, from ricochet, impact with a hard target, or careless handling in evidence collection, can degrade or destroy the marks entirely.
Class, Individual, and Subclass Characteristics
Not every mark on a fired bullet or casing points to a single gun. Examiners divide the marks into categories that carry very different weight.
Class characteristics are design-level features shared by every firearm of the same make and model. The number of rifling grooves, their direction of twist, and the width of the land impressions between them are class features. They can narrow the field to a group of firearms but cannot identify one specific weapon.
Individual characteristics are the random microscopic imperfections unique to a single barrel, firing pin, or breech face. Striation marks on a fired bullet are the classic example. Because these irregularities arise from incidental manufacturing variation and wear, no two firearms are expected to produce identical individual marks, even guns that came off the same assembly line on the same day.1National Institute of Justice. Firearms Examiner Training – Class and Individual Characteristics
Between those two sits a third and more dangerous category. Subclass characteristics appear on firearms manufactured consecutively with the same tooling. Because the tooling wears gradually, a batch of barrels or breech faces produced in sequence can share surface features that look like individual marks but actually belong to a small group of weapons. An examiner who mistakes subclass marks for unique identifiers can wrongly link a casing to the wrong gun. Training materials instruct examiners to research the manufacturing processes for the specific make and model and to consult published literature when subclass influence is possible.2National Institute of Justice. Firearms Examiner Training – Physical Characteristics
From Crime Scene to Comparison Microscope
Comparison requires two things: the evidence recovered from the scene, and a known reference sample from the suspect firearm. To produce the reference, the laboratory test-fires the weapon into a controlled recovery system that captures the bullet without damaging its surface. The standard method is a water recovery tank, where the bullet is fired at a downward angle into water that decelerates it gently. Commercial water tanks became the norm in the mid-1980s. Some laboratories still use cotton-filled boxes, though those carry a risk of gunshot residue igniting the material.3National Institute of Justice. Firearms Examiner Training – Bullet Recovery Equipment
The core comparison happens under a comparison microscope, an instrument that pairs two independent microscopes through an optical bridge so the examiner sees both specimens side by side in a single field of view. The examiner rotates and aligns the two samples, looking for agreement in the individual characteristics that goes beyond what would be expected between marks made by different tools.
Some laboratories now supplement or replace this with three-dimensional surface scanning. These systems capture the topography of a casing or bullet in high resolution, producing digital models that examiners can rotate, annotate, and share electronically. Research indicates that high-quality 3D examination takes less time and produces more accurate conclusions than traditional microscopy, while also creating a documented record of exactly which regions of the surface the examiner relied on.4Office of Justice Programs. Evaluation of 3D Virtual Comparison Microscopy for Firearm Forensics Within the Crime Lab
The Four Possible Conclusions
The Association of Firearm and Toolmark Examiners defines four outcomes for a comparison. An identification means the agreement of individual and class characteristics exceeds what could occur between marks made by different tools. An elimination means significant disagreement. An inconclusive result means there is some agreement but not enough for identification, or some disagreement but not enough for elimination. Unsuitable for comparison applies to evidence too damaged or mutilated to bear useful marks.5National Institute of Justice. Firearms Examiner Training – AFTE Theory of Identification
The framework is important to understand because it drives how error rates get reported. Whether inconclusive results count as errors, non-errors, or something in between changes accuracy statistics dramatically, as discussed below.
How NIBIN Connects Cases Across Jurisdictions
Criminals frequently use the same gun in multiple crimes, and those crimes often cross city and state lines. To surface connections that local investigators would miss on their own, the Bureau of Alcohol, Tobacco, Firearms and Explosives established the National Integrated Ballistic Information Network in 1997. NIBIN stores high-resolution digital images of spent cartridge casings recovered from crime scenes and test fires, and makes them searchable by law enforcement nationwide. ATF operates the infrastructure at no cost to partner agencies.6Bureau of Alcohol, Tobacco, Firearms and Explosives. National Integrated Ballistic Information Network Fact Sheet
As of fiscal year 2024, NIBIN had 378 partner locations and more than 7 million pieces of ballistic evidence in the database, generating over 217,000 investigative leads in a single year. The system is used exclusively for criminal investigations and does not capture ballistic data at the point of manufacture, importation, or sale.7Bureau of Alcohol, Tobacco, Firearms and Explosives. National Integrated Ballistic Information Network (NIBIN)
Technicians acquire images using the BrassTrax acquisition station, which uses topographic scanning to render a three-dimensional image of the casing surface.8National Institute of Justice. Assessing the Quality of 3-Dimensional Imaging on the BrassTrax HD3D System Correlation software then ranks entries by visual similarity and produces a list of potential associations.
A NIBIN lead is not a match. It is an unconfirmed potential association identified by the software and a trained technician, meaning the digital images look similar enough to justify further work. Converting a lead into a confirmed NIBIN hit requires a firearms examiner to physically compare the actual casings under a comparison microscope. Only after that hands-on examination does the association become a verified forensic link.9Bureau of Alcohol, Tobacco, Firearms and Explosives. National Integrated Ballistic Information Network A cold hit occurs when the system links evidence from a case to a casing from an entirely unrelated one that investigators had no reason to connect. The median time between a shooting and a confirmed hit is roughly 101 days, though the fastest arrive within a week.10Office of Justice Programs. A Descriptive Process and Outcome Evaluation of the Use of NIBIN
How Reliable Is Ballistic Fingerprinting?
The entire discipline rests on one assumption: that every firearm leaves marks unique enough to be distinguished from every other firearm. Two major government-commissioned reviews have questioned whether the science adequately supports that assumption.
In 2009, the National Academy of Sciences published a sweeping assessment of forensic disciplines and concluded that the validity of the fundamental assumptions of uniqueness and reproducibility “has not yet been fully demonstrated.” The report found that examiners’ decisions remain subjective and based on unarticulated standards, that the field lacks a statistical foundation for estimating error rates, and that no one could specify how many points of similarity are necessary for a given level of confidence.11Office of Justice Programs. Strengthening Forensic Science in the United States – A Path Forward
Seven years later, the President’s Council of Advisors on Science and Technology reached a similar verdict. PCAST concluded that firearms analysis “currently falls short of the criteria for foundational validity” because only a single properly designed study had attempted to measure accuracy. That study, conducted by the Ames Laboratory for the Department of Defense, estimated a false positive rate of 1 in 66, with a confidence bound suggesting the rate could be as high as 1 in 46. PCAST stated that claims of higher accuracy were “not scientifically justified at present” and recommended that the field move toward objective, algorithm-based image comparison rather than relying on subjective examiner judgment.12The White House (Obama Administration Archives). Report to the President – Forensic Science in Criminal Courts
Reported error rates also vary sharply depending on how laboratories count inconclusive results. When examiners who cannot make a call report “inconclusive” and those results are set aside, published studies show error rates below 1%. A 2022 peer-reviewed analysis illustrated the problem: in one major study, the false positive rate came in at 0.7% when inconclusives were treated as neutral non-errors but jumped to 66% when inconclusives on different-source evidence were treated as potential errors. Another study showed an overall error rate of 1.25% with no inconclusives, rising to 31.8% when examiner inconclusives were counted as errors.13PMC (PubMed Central). Inconclusives, Errors, and Error Rates in Forensic Firearms Analysis – Three Statistical Perspectives The point is not that firearms analysis is unreliable, but that accuracy claims depend heavily on statistical choices that are not settled.
What Examiners Can and Cannot Say in Court
The days when a firearms examiner could tell a jury a bullet came from one specific gun “to the exclusion of all other firearms in the world” are largely gone. Under Department of Justice Uniform Language guidelines issued in 2020, DOJ examiners cannot assert that a match is based on the “uniqueness” of the evidence, cannot claim a zero error rate, and cannot use terms like “100% certainty” or “reasonable degree of scientific certainty” unless a judge specifically orders it.14U.S. Department of Justice. Approved Uniform Language for Testimony and Reports for the Forensic Firearms/Toolmarks Discipline
Federal courts have imposed their own limits through case law. In United States v. Blackman, the court allowed ballistic experts to testify but barred them from claiming 100% certainty or using language implying that firearms analysis is an exact science. The court cited earlier decisions reaching similar results: one permitted only a “more likely than not” standard, another allowed testimony to a “reasonable degree of ballistic certainty” but not absolute certainty, and another restricted the expert to stating that the suspect firearm “cannot be excluded” as the source.15CaseMine. United States v Blackman
Behind those limits sits Federal Rule of Evidence 702, which as amended in December 2023 requires the party offering expert testimony to show it is “more likely than not” that the testimony rests on sufficient facts, reliable methods, and a reliable application of those methods.16U.S. Courts. Federal Rules of Evidence Pamphlet – Rule 702 Trial judges evaluate reliability using the Daubert standard from Daubert v. Merrell Dow Pharmaceuticals, which asks whether a technique can be tested, whether it has been peer-reviewed, its known or potential error rate, and whether it has gained general acceptance in the scientific community.17Justia. Daubert v Merrell Dow Pharmaceuticals Inc – 509 US 579 Most courts have admitted ballistic testimony under Daubert, but admission increasingly comes with conditions on how confidently the examiner can phrase the conclusion in front of a jury.
So a modern jury is more likely to hear that evidence is “consistent with” having been fired from a particular weapon, or that the weapon is “identified as the source” under the AFTE framework, than the sweeping declarations of certainty that once defined firearms testimony in federal court.