DNA evidence in criminal cases is biological material — blood, semen, saliva, skin cells, hair with a root — that a lab can tie to a specific person through genetic profiling, and whether it convicts, exonerates, or gets thrown out depends on how it was collected, tested, matched, and challenged. Every person other than an identical twin carries a unique genetic code, and forensic labs use that code to include or exclude someone as the source of a sample. The national DNA database has assisted in more than 758,000 investigations, and DNA testing has also freed over 375 people who were wrongfully convicted. Both numbers matter, because the same technology cuts in both directions.
What Counts as DNA Evidence
Blood and semen are the strongest sources because they deposit large quantities of DNA in concentrated stains, often on clothing, bedding, flooring, or surfaces near a point of physical contact. Saliva is another reliable source, recovered from drinking glasses, cigarette butts, envelope seals, and bite marks. Hair can yield a full profile, but only if the root is attached; a bare shaft lacks the nuclear DNA a standard test needs.
Then there is touch DNA — the microscopic skin cells someone leaves when they handle an object. Technicians swab door handles, weapon grips, steering wheels, and shirt collars for these traces. Modern profiling is sensitive enough to build a result from very little material, and that sensitivity creates a real problem: a handshake can move one person’s cells onto another person’s hands, and that second person can then deposit the first person’s DNA on an object they touch later. This is called secondary transfer, and it means a touch DNA hit does not automatically prove that the person named in the profile ever touched the evidence.
How Investigators Collect Samples
At a scene, technicians lift biological fluids with sterile cotton swabs. They also seize whole items — stained garments, discarded weapons — and package them in breathable paper bags rather than plastic, because trapped moisture grows mold that degrades DNA. Gloves and masks keep the officers’ own genetic material out of the evidence, and each item is photographed in place before it is moved.
Buccal Swabs and Warrants
To compare a scene sample against a suspect, the lab needs a known reference. The standard method is a buccal swab, a sterile applicator rubbed against the inside of the cheek. In Maryland v. King (2013), the Supreme Court held that taking a cheek swab from someone arrested for a serious offense and brought in for booking is reasonable under the Fourth Amendment, comparable to fingerprinting.1Justia Law. Maryland v. King 569 U.S. 435 (2013) For a suspect who has not been arrested, a search warrant supported by probable cause is generally required.2National Policing Institute. Fourth Amendment Limitations on DNA Collection, Procurement, and Retention
Police also collect abandoned DNA from items a person discards in public, like a coffee cup or a used tissue. Courts have generally held that discarding an item in a public space surrenders any reasonable expectation of privacy in it, so no warrant is required.3The Heritage Foundation. GEDMatch and the Fourth Amendment – No Warrant Required
Rapid DNA at Booking
Rapid DNA machines generate a profile from a cheek swab in roughly 90 minutes at the booking station, without sending the sample to an outside lab. The Rapid DNA Act of 2017 gave the FBI Director authority to set standards for these devices, and approved units can search CODIS directly during booking. The agency’s state must have a law authorizing DNA collection at arrest, and the booking station must integrate the device with its electronic fingerprint system and the state identification bureau’s network.4FBI.gov. Guide to All Things Rapid DNA A database hit can now surface before the suspect’s first court appearance.
How the Lab Builds a Profile
Once a sample reaches the lab, technicians break down cell membranes to separate the DNA from other cellular debris and then measure how much human DNA the extract contains. If the quantity is too low or the DNA is too degraded, standard methods may fail.
STR Analysis
The current standard is Short Tandem Repeat analysis, which examines specific locations on the DNA strand where short chemical sequences repeat a different number of times in different people. Before analysis, technicians use Polymerase Chain Reaction to copy the target regions millions of times, which makes small or partially degraded samples workable. Since January 2017, labs participating in the national database have been required to type a minimum of 20 core locations, up from 13, which sharply reduced the chance of a coincidental match between unrelated people.5Office of Justice Programs. Switching to 20 Core CODIS Loci and the Impact on SAKI Testing The output is a numerical profile that can be compared across databases without the physical tissue.
Mitochondrial DNA
When nuclear DNA is too degraded or absent, such as skeletal remains exposed to the elements or hair fragments without roots, the lab can turn to mitochondrial DNA. There is more of it per cell, so it survives damage better. It is particularly useful for rootless hairs and small hair fragments under one centimeter long, which would be useless for standard STR testing.6National Center for Biotechnology Information. Mitochondrial DNA Analysis of 114 Hairs Measuring Less Than 1 cm From a 19-Year-Old Homicide The tradeoff: mitochondrial DNA is inherited maternally, so it cannot distinguish between individuals in the same maternal line. It narrows the field rather than pinpointing one person.
Mixed Profiles
Crime scene samples often contain DNA from more than one contributor, and older interpretation methods handled mixtures poorly. Labs now use probabilistic genotyping software that applies statistical models to estimate how likely it is that a particular person contributed to a mixed profile. The output is a likelihood ratio: how many times more probable the evidence is if the suspect contributed than if a random, unrelated person did.7National Institute of Standards and Technology. DNA Mixture Interpretation – A NIST Scientific Foundation Review Defense attorneys increasingly challenge these programs, particularly when the underlying source code is proprietary and not open to independent review.
How Databases Produce Leads
Forensic profiles are uploaded to the Combined DNA Index System, or CODIS, a tiered platform that connects local, state, and national laboratories. The top tier is the National DNA Index System, which allows searches across the country.8Federal Bureau of Investigation. CODIS and NDIS Fact Sheet As of November 2025, NDIS holds more than 19.2 million offender profiles, 6.1 million arrestee profiles, and 1.4 million forensic profiles, and it has produced over 781,000 hits assisting in more than 758,000 investigations.9FBI.gov. CODIS-NDIS Statistics
Federal and state statutes govern who has to submit a sample. Most states require collection from anyone convicted of a felony, and many also authorize collection from people arrested for certain serious offenses before conviction.8Federal Bureau of Investigation. CODIS and NDIS Fact Sheet When a scene profile matches a stored one, the result is called a cold hit — a lead in a case that had no suspect through conventional work.
Familial Searching
When CODIS returns no exact match, some jurisdictions run a familial search to identify a close biological relative of the unknown source. Because parents pass alleles to children in predictable patterns, close relatives share more markers than unrelated people. A familial search flags partial matches and uses mathematical modeling to estimate whether the similarity likely reflects kinship rather than coincidence.10Bureau of Justice Assistance. An Introduction to Familial DNA Searching The result is an investigative lead, not a direct identification.
Forensic Genetic Genealogy
A newer technique goes outside CODIS entirely. Forensic genetic genealogy involves uploading crime scene DNA to consumer genealogy databases to find distant relatives of an unknown source, then building family trees to close in on a suspect. The Department of Justice’s interim policy restricts this technique to unsolved violent crimes, primarily homicides and sexual offenses, and only after CODIS has failed to produce a match. The investigating agency must also consult with the CODIS laboratory and a prosecutor before proceeding.11U.S. Department of Justice. Forensic Genetic Genealogical DNA Analysis and Searching The method has cracked decades-old cold cases and has drawn steady privacy criticism, because it exposes the genetic information of relatives who never consented to law enforcement access.
Getting DNA Evidence Into Trial
A match only matters if it survives legal scrutiny. Three doctrines do most of the work.
Chain of Custody
The prosecution has to produce an unbroken record of who handled the evidence from collection through trial.12National Institute of Justice. What Every Law Enforcement Officer Should Know About DNA Evidence – Chain of Custody Those records have to show who collected each sample, how it was packaged, sealed, and stored, and every transfer between people or facilities.13National Institute of Justice. Chain of Custody Record An unexplained gap opens the door to arguments about contamination or tampering, and a judge can suppress the evidence.
Frye and Daubert
Two tests decide whether forensic science is reliable enough for a jury to hear. Under the Frye standard, from the 1923 case Frye v. United States, the technique must be generally accepted in the relevant scientific community.14Legal Information Institute. Frye Standard The federal courts and most states now use the Daubert standard from Daubert v. Merrell Dow Pharmaceuticals, under which the trial judge acts as gatekeeper and weighs factors including whether the technique has been tested, whether it has been peer reviewed, its known error rate, and whether it follows maintained standards.15Legal Information Institute. Daubert v. Merrell Dow Pharmaceuticals, Inc. Defense lawyers commonly target the specific lab’s protocols, the analyst’s qualifications, or the amplification error rate.
Confronting the Analyst
Even when the science holds up, the Sixth Amendment’s Confrontation Clause requires the prosecution to put the analyst who did the testing on the stand. In Melendez-Diaz v. Massachusetts (2009), the Supreme Court held that forensic lab reports are testimonial, so the defendant has the right to confront the analyst who prepared them rather than have the report read into evidence.16Justia Law. Melendez-Diaz v. Massachusetts 557 U.S. 305 (2009) Two years later, in Bullcoming v. New Mexico, the Court held that a surrogate analyst who did not personally perform or observe the testing cannot stand in.17Justia Law. Bullcoming v. New Mexico 564 U.S. 647 (2011) If the original analyst has left the agency, the prosecution has a real problem getting the evidence to the jury.
How the Match Number Sounds
Labs express the strength of a match as a likelihood ratio comparing two scenarios: the probability of the evidence if the suspect is the contributor, versus the probability if a random, unrelated person is.18National Institute of Justice. Likelihood Ratio – Mixtures An expert might testify that the evidence is “10 billion times more likely if the suspect contributed than if an unknown, unrelated person did.” In clean single-source samples, those numbers hold up. In mixed or degraded samples, the ratios drop and the interpretation becomes contested, which is where most defense challenges focus.
Common Defense Challenges
DNA evidence carries heavy weight with juries, and that weight becomes a liability when the evidence is weaker than it looks.
Secondary and tertiary transfer is one of the most underappreciated issues. Cells travel through handshakes, shared objects, and laundry, and modern profiling can produce a full result from very little material. A person’s DNA on a weapon does not necessarily mean they touched it. Defense teams challenge touch DNA results by presenting transfer studies showing how easily DNA moves between surfaces and people.
Lab contamination is another recurring risk. Even under strict protocols, a single error — an uncapped tube, a glove not changed between samples, a shared workspace — can introduce DNA that has nothing to do with the crime. Chain-of-custody gaps make it harder to pinpoint where such contamination could have crept in.
Mixed profiles bring their own difficulties. When three or four people have contributed to a sample, even probabilistic genotyping software depends on assumptions about the number of contributors and the extent of degradation. Different packages can produce different likelihood ratios from the same raw data, and defense experts increasingly demand access to the source code to test whether the algorithms perform as claimed.
Finally, a match shows biological material was present. It does not explain how or when. A suspect’s DNA in a victim’s apartment means far less if the suspect is a coworker who visited the week before. Context and corroboration still do work that the profile alone cannot.
Post-Conviction DNA Testing
The same technology that convicts also undoes wrongful convictions. More than 375 people in the United States have been exonerated through DNA testing after serving an average of 14 years in prison.
The Federal Right to Testing
Under 18 U.S.C. § 3600, a person sentenced to imprisonment or death for a federal offense can file a motion asking the court to order DNA testing of specific evidence. The court must grant the motion if the applicant meets every requirement: asserting actual innocence under penalty of perjury, showing the evidence was secured during the original investigation, showing the perpetrator’s identity was at issue at trial, and showing that the proposed testing could produce new material evidence raising a reasonable probability the applicant did not commit the offense.19Office of the Law Revision Counsel. 18 U.S. Code 3600 – DNA Testing If the evidence was tested before, retesting is available only with a newer method substantially more probative than the earlier one. Congress and all 50 states have some form of post-conviction DNA access statute, but the eligibility rules vary widely.
Preservation of the Evidence
A right to test evidence is meaningless if the evidence is gone. Under 18 U.S.C. § 3600A, the federal government must preserve biological evidence from any federal case that resulted in a prison sentence. The obligation runs until the defendant has exhausted direct appeals, been notified that the evidence may be destroyed, and failed to file a motion for testing within 180 days of that notice.20Office of the Law Revision Counsel. 18 U.S. Code 3600A – Preservation of Biological Evidence State laws vary. Destruction before a defendant has a real chance to request testing remains one of the biggest practical barriers to post-conviction relief.
Removing a DNA Profile From CODIS
If charges are dismissed, a case ends in acquittal, or a conviction is overturned, the person may be entitled to have their profile expunged from the national database and the physical sample destroyed. For federal arrests and convictions, the request goes in writing to the FBI’s Federal DNA Database Unit with a certified copy of a final court order documenting the dismissal, acquittal, or overturned conviction. The order must be signed by a judge, dated, and contain enough identifying information to link the request to the correct record — at minimum the person’s full name and either a Social Security number or date of birth. Without a certified order, the FBI will not process the request.21Federal Bureau of Investigation. DNA Fingerprint Act of 2005 Expungement Policy
This federal process reaches only records from federal arrests or convictions. State and local records follow separate state procedures with their own requirements and timelines. Anyone seeking removal of a state-level profile should check with the state crime lab or the attorney who handled the case, because the window to request expungement often starts running at disposition.