Genetic genealogy helps solve crimes by taking DNA left at a crime scene, running it through a public ancestry-style database to find the perpetrator’s distant relatives, and then building family trees outward from those relatives until investigators can identify one likely suspect. The suspect’s own DNA does not need to be in any police database. A third or fourth cousin’s DNA, uploaded years earlier for personal ancestry research, is enough of a thread to pull.
The technique has been credited with resolving more than 545 cases by the end of 2022, covering both criminal identifications and unidentified human remains, and the count has climbed since as more agencies have adopted it.
Why Police Needed Something Beyond CODIS
The FBI’s Combined DNA Index System, or CODIS, has been the workhorse of forensic DNA for decades. It compares profiles built from short tandem repeats (STRs) at 20 specific spots in a person’s DNA.1Federal Bureau of Investigation. CODIS and NDIS Fact Sheet Twenty markers are enough to confirm or exclude a direct match, but they carry too little information to reveal distant relationships. A fourth cousin looks like a stranger in CODIS.
CODIS has a second, structural limit. It only holds profiles from people who have been arrested, convicted, or otherwise legally required to submit a sample. If the perpetrator has never been in the system, a CODIS search returns nothing.
Genetic genealogy works around both limits. Instead of 20 STR markers, it reads hundreds of thousands of single nucleotide polymorphisms (SNPs) scattered across the genome. With that much more data, two people who share a great-great-great-grandparent still share detectable stretches of matching DNA. And the databases being searched are not police databases at all. They are filled with ordinary people who uploaded their own DNA for ancestry research.
The Four Steps From Crime Scene to Suspect
Building an SNP Profile From the Evidence
The process starts with biological evidence: blood, semen, saliva, skin cells, or similar material recovered from a crime scene. A specialized vendor laboratory extracts DNA from the sample and runs it through a microarray chip that reads hundreds of thousands of SNP positions across the genome. The output is a data file formatted for genealogy databases, not the STR profile stored in CODIS.
This step needs a reasonably intact sample. Badly degraded evidence sometimes cannot produce a usable SNP profile, though laboratory techniques for compromised samples keep improving. Semen from sexual assault cases tends to yield the highest-quality DNA. Trace evidence, like touch DNA from a doorknob, is far less reliable.2National Institutes of Health. Why We Fear Genetic Informants
Uploading to a Genealogy Database
The SNP profile is then uploaded to a genealogy database that permits law enforcement searches. The database’s software compares the profile against every opted-in user and returns a list of matches ranked by how much DNA each match shares with the unknown person. More shared DNA means a closer relative. A large match might indicate a second cousin, a smaller one a fourth cousin. Even distant matches are useful, because each match locates the unknown person on some branch of a family tree.
Building Family Trees
This is the most labor-intensive part, and it is where old-fashioned genealogy takes over. Professional genealogists take each DNA match and research that person’s ancestry backward through generations using birth and death records, marriage records, census data, obituaries, and newspaper archives. The goal is to find common ancestors shared by multiple matches, then trace forward through every descendant until the trees converge on a small group of living people who could be the source of the crime scene DNA.
The trees can span five or six generations and involve thousands of people. Genealogists cross-reference the emerging pool against case facts: the geographic area of the crime, the approximate age of the suspect, physical descriptions, and any other detail that can eliminate branches. The work can take weeks or months. When it succeeds, it usually narrows the field to one person or a handful.
Confirming the Match With a Direct Sample
Genetic genealogy produces an investigative lead, not a conviction. The final step is always a traditional forensic comparison using the suspect’s own DNA. Officers often collect a discarded sample, such as a cup, utensil, or cigarette butt left in a public place.3National Institute of Justice. DNA – A Prosecutors Practice Notebook – Abandoned Sample They may also obtain a warrant for a direct sample. That sample is then analyzed using standard STR profiling and compared to the crime scene evidence. A match here provides the identification used at trial.
The 2018 arrest of Joseph James DeAngelo Jr. as the suspected Golden State Killer, tied to at least 13 murders and more than 50 sexual assaults between 1974 and 1986, is the case that put this workflow on the map. DeAngelo had never been required to give a DNA sample, so CODIS had nothing on him. A GEDmatch upload, family-tree work, and a discarded-sample confirmation closed the case. He later pleaded guilty and was sentenced to life in prison.
Which Databases Police Can Actually Use
Not every consumer DNA service is open to investigators, and this is one of the most misunderstood parts of the process. Two platforms currently permit law enforcement uploads: GEDmatch and FamilyTreeDNA. The major consumer testing companies do not.
GEDmatch is a free platform where users upload raw DNA files from any testing company. Each user chooses a setting when they upload. Opting in allows the profile to be compared against law enforcement submissions for violent crime investigations and human remains identification. Opting out blocks matching for violent crime kits but still permits matching for unidentified remains cases. Police must use a dedicated portal called GEDmatch PRO, and searches are limited to violent crimes defined as murder, manslaughter, aggravated rape, robbery, or aggravated assault.4GEDmatch. Terms of Service
FamilyTreeDNA operates a similar framework. Agencies must register a dedicated account, submit documentation, and receive approval before uploading forensic samples. Searches are limited to identifying deceased persons and identifying suspects in violent crimes, which the company defines to include homicide, sex crimes, armed robbery, aggravated assault, and other violent offenses recognized under the FBI’s Uniform Crime Reporting program.5FamilyTreeDNA. Law Enforcement Guide
23andMe and AncestryDNA do not allow investigators to upload a forensic profile or run genetic-match searches. 23andMe requires a valid court order, subpoena, or search warrant before it will consider producing customer information, and has stated that it has never voluntarily released customer data to law enforcement.623andMe Customer Care. How 23andMe Responds to Law Enforcement Requests for Customer Information AncestryDNA has a similar policy.
When Police Are Allowed to Use the Technique
The Department of Justice issued an interim policy in 2019 that governs federal investigations and has served as a model for many state and local agencies. It restricts forensic genetic genealogy to unsolved violent crimes and unidentified human remains believed to be homicide victims. “Violent crime” means homicide or sexual offenses, though a prosecutor can authorize use for other violent crimes such as robbery or aggravated assault when there is an ongoing public safety threat or national security concern.7United States Department of Justice. Interim Policy Forensic Genetic Genealogical DNA Analysis and Searching
Investigators cannot go straight to a genealogy database. The policy requires that the crime scene profile first be uploaded to CODIS and that CODIS searches fail to produce a confirmed match. Agencies must also have pursued reasonable traditional leads and, where applicable, entered the case into the National Missing and Unidentified Persons System (NamUs) and the Violent Criminal Apprehension Program (ViCAP) database. A prosecutor must sign off before the process begins.7United States Department of Justice. Interim Policy Forensic Genetic Genealogical DNA Analysis and Searching
In practice, this is a tool of last resort. It is aimed at cold cases and the most serious offenses. It is not used to identify shoplifters or in ordinary property crimes.
Identifying Unknown Human Remains
The same workflow is used to put names on unidentified dead. Thousands of “John Doe” and “Jane Doe” cases sit open across the country because no matching missing person report was ever filed, or because the remains were too decomposed for visual identification. Genealogy-database matching connects those remains to a family tree and, often, to living relatives.
The DOJ interim policy explicitly covers unidentified remains cases, and GEDmatch’s terms allow matching against law enforcement kits for remains identification even for users who have opted out of violent crime searches.4GEDmatch. Terms of Service
Where the Technique Runs Into Trouble
Genetic genealogy is not a guaranteed answer. Cases stall for a few recurring reasons:
- DNA quality. Old or environmentally degraded evidence may not produce a strong enough SNP profile, and weak profiles produce incomplete or unreliable matches.
- Database demographics. The databases open to law enforcement are disproportionately populated by people of European descent. Cases involving suspects from other backgrounds tend to produce fewer and more distant matches, which makes tree-building far harder and sometimes impossible.
- Time and cost. Generating an SNP profile has become cheaper, but the genealogy research itself is skilled, slow, and labor-intensive. Many agencies do not have the budget or trained staff to run these cases routinely.
- Gaps in public records. Adoptions, name changes, incomplete vital records, and recent immigrants with limited U.S. documentation all create holes that can stall an investigation for good.
The Privacy Trade-Off
Uploading DNA to a genealogy database shares information about the uploader’s relatives too, whether those relatives consented or not. By one estimate, law enforcement access to DNA from as little as two percent of the U.S. population could make roughly 90 percent of Americans of European descent genetically identifiable, including people who have never taken a DNA test.2National Institutes of Health. Why We Fear Genetic Informants
The Supreme Court has not directly addressed forensic genetic genealogy. In Maryland v. King (2013), the Court upheld DNA cheek swabs during arrest bookings, reasoning that CODIS loci do not reveal private medical traits, but it flagged that analyzing DNA for “hereditary factors not relevant to identity” would raise additional privacy concerns.8Legal Information Institute. Maryland v King Genetic genealogy involves far more of the genome than CODIS. Legal scholars have also argued that Carpenter v. United States (2018), which required a warrant for extensive digital records held by a third party, applies here, because a person shares their DNA with relatives involuntarily and cannot sever that connection.9National Institutes of Health. Familial Searches, the Fourth Amendment, and Forensic Genetic Genealogy
A handful of states have enacted laws that specifically regulate the technique, some requiring judicial authorization before an upload, some adding protections for non-suspect third parties whose DNA turns up in results, and some letting defendants use the same technique to prove their innocence. Most jurisdictions still operate under the DOJ’s voluntary interim policy or have no specific rules at all.