GEDmatch results are organized around a few core pages, and the most important number on any of them is the total centimorgans (cM) you share with a match. That single figure, combined with the number and size of matching DNA segments, tells you roughly how closely related you are to another person in the database. But the platform offers more than a ranked list of cousins. Admixture calculators, chromosome browsers, and phased kit tools each answer different questions about your ancestry, and misreading any one of them can send your genealogical research sideways.
The One-to-Many Page
When you first run a comparison on GEDmatch, the “One-to-Many” results page is what you see. It lists every person in the database whose DNA overlaps with yours above a minimum threshold, ranked from highest shared DNA to lowest. Each row shows a kit number (the anonymous identifier for that person’s upload), the total centimorgans shared, the number of shared segments, and the size of the largest segment. Some rows also display the person’s reported ancestry or a user-chosen alias, though many kits are anonymous or pseudonymous.
Your closest genetic relatives appear at the top. A parent or sibling will share thousands of centimorgans. A first cousin shares somewhere around 850 cM on average, though individual variation is wide. As you scroll down the list, total shared cM drops, and the matches become more distant. By the time you reach people sharing 20 or 30 cM, you are looking at fourth or fifth cousins, and the genealogical connection could be anywhere in the last two centuries of your family tree. Research on large genetic genealogy databases has found that roughly 60 percent of searches return at least one match sharing 100 cM or more, which typically corresponds to a third-cousin-level relationship or closer.1Forensic Science International: Genetics. An empirical investigation into the effectiveness of genetic genealogy to identify individuals in the UK
Two columns deserve special attention even on this overview page. The “largest segment” column tells you the size of the single longest stretch of DNA you share with that match. A person who shares 90 cM total but whose largest segment is only 10 cM may be related to you through multiple very distant lines rather than one closer relationship. In contrast, someone sharing 90 cM with a largest segment of 45 cM probably connects to you through a single recent ancestor. The largest segment is often a better clue to genuine recent kinship than the total cM alone.
What Centimorgans Actually Tell You
A centimorgan is a unit of genetic distance, not a fixed length of physical DNA. It measures the probability that a stretch of a chromosome gets shuffled during the process that creates eggs and sperm. Because shuffling (recombination) happens somewhat randomly, the amount of DNA any two relatives share varies, sometimes by a lot. Two half-siblings might share 1,600 cM or 1,900 cM. Two second cousins might share anywhere from roughly 40 to 400 cM. GEDmatch gives you the measured number. Translating it into a specific relationship is where things get tricky.
Online tools like the Shared cM Project (a crowd-sourced dataset of known relationships and their measured cM values) can help you see the range of possibilities. If you share 200 cM with someone, that is consistent with a second cousin, a half first cousin once removed, or a great-great-grandparent’s other descendant through a different branch. The cM figure narrows the field but rarely points to a single relationship. You need traditional genealogy, family trees, and sometimes additional DNA matches to pin it down.
A common mistake is treating the cM number as an exact measurement rather than an estimate. GEDmatch applies a minimum segment threshold (usually 7 cM by default) below which small matching chunks are ignored. Those tiny segments are more likely to be “identical by state” rather than “identical by descent,” meaning they look similar by coincidence, not because you inherited them from a common ancestor. Adjusting that threshold up or down changes your total shared cM, so two people running the same comparison with different settings can get different numbers.
One-to-One Comparisons and the Chromosome Browser
The One-to-Many page is a survey. The One-to-One comparison is a deep dive. When you plug in your kit number and a specific match’s kit number, GEDmatch shows you exactly where on each chromosome your DNA overlaps. The result comes both as a table (listing each chromosome, the start and stop positions of each matching segment, and the segment’s size in cM) and as a visual chromosome browser, where colored bars mark the regions you share.
The chromosome browser is useful for a technique called “triangulation.” If you and two other matches all share a segment at the same location on the same chromosome, all three of you probably inherited that stretch of DNA from the same ancestor. Triangulated groups help you assign matches to specific branches of your family tree, maternal or paternal side, and sometimes to a specific grandparent line.
One thing worth knowing is that GEDmatch’s chromosome browser shows half-identical regions (HIR) by default, meaning you match on one copy of a chromosome. Fully identical regions, where you match on both copies, show up only between close relatives like full siblings. If you are comparing yourself with someone and the browser shows large segments of full identity across many chromosomes, that signals a close relationship on both the maternal and paternal sides, which can happen when relatives intermarry.
Making Sense of Admixture Calculators
GEDmatch hosts a battery of admixture calculators with names like MDLP, Eurogenes, Dodecad, and HarappaWorld. Each one compares your DNA to a set of reference populations and estimates what percentage of your ancestry comes from various geographic or ethnic groups. This is where people most often get confused, because different calculators give different results for the same person.
The differences are not errors. Each calculator was built by a different researcher or hobbyist, using different reference populations and different statistical methods. A calculator built to distinguish fine-grained European sub-populations will lump all East Asian ancestry into a single category, while one designed for South Asian detail may be crude about Scandinavian versus Mediterranean ancestry. No single calculator is “right.” They are lenses, each optimized for a different part of the world.
If your ancestry is mostly from one region, pick the calculator designed for that region and treat its estimates as rough guides. If you are of mixed background, try several calculators and look for consistencies. A result that says you are 8 percent “West African” on one model and 10 percent on another is telling you something real. A result that says you are 3 percent “Oceanian” on one model and zero on all others is probably noise. The Oracle tool, which appears alongside the admixture output, tries to identify which modern populations your admixture pattern most closely resembles. It is fun to play with but should be treated as a conversation starter, not a definitive answer about your ethnic identity.
Phased Kits and Reconstructed Results
Your DNA comes in pairs, one set from each parent, but most consumer DNA tests read both copies at once without distinguishing which came from which parent. “Phasing” is the process of separating those two copies. GEDmatch allows users to create phased kits when both a parent and child have tested. The parent’s data helps the software figure out which of the child’s DNA variants came from that parent and which came from the other.
Phased kits are valuable because they cut down on false matches. Without phasing, GEDmatch might report a match that looks legitimate but actually stitches together a segment from your maternal chromosome and a segment from your paternal chromosome. These composite or “pseudo” segments inflate your shared cM total and can mislead you into thinking a relationship is closer than it is. When you compare using a phased kit, only segments genuinely inherited from one parent at a time are counted. Research using GEDmatch with family-based phasing and identical-by-descent segment sharing has been applied in population studies to identify distant relatives across continents, including work connecting members of the African diaspora with relatives in Ghana.2bioRxiv. Genetic Genealogical Methods Used to Identify African American Diaspora Relatives in the Study of Family Identity among Ghanaian Members of the Kassena Ethnic Group
GEDmatch also supports “reconstructed” kits for deceased or untested relatives. If multiple children of an untested parent have all uploaded their DNA, the platform can partially reconstruct the parent’s genotype by finding what the children have in common. Reconstructed kits are inherently less complete than real test results, but they can still surface matches that none of the children would find individually, because each child inherited only about half of the parent’s DNA.
When Shared DNA Overstates a Relationship
Endogamy is the single biggest source of confusion for people trying to interpret GEDmatch results. In populations where people married within a relatively small community for generations, like Ashkenazi Jewish, Acadian, or certain island populations, everyone in the group shares more DNA with everyone else than the standard cM-to-relationship charts predict. Two Ashkenazi Jewish people who share 100 cM might actually be fifth or sixth cousins rather than the third cousins that 100 cM normally suggests.
GEDmatch does not automatically adjust for endogamy. The cM numbers it reports are what they are. If you come from an endogamous background, you will see inflated totals across the board, and you will have far more matches than the average user. The practical workaround is to pay less attention to total cM and more to the largest segment. In endogamous relationships, the shared DNA tends to be spread across many small segments rather than concentrated in a few large ones. A match sharing 100 cM whose largest segment is 15 cM is probably more distant than a match sharing 100 cM whose largest segment is 40 cM.
Pedigree collapse, where your ancestors are related to each other on paper in more than one way, has a similar effect even outside classically endogamous groups. If your great-great-grandparents on different branches happened to be cousins, you carry more DNA from that shared lineage than the standard charts account for. This is more common than most people realize, especially in rural communities or among families who stayed in the same region for many generations. The result is the same: GEDmatch will overestimate closeness.
Segment Size Thresholds and Noise
GEDmatch lets you adjust the minimum segment size considered a match. The default is usually 7 cM, but you can raise or lower it. Lowering the threshold to 5 or even 3 cM pulls in more matches, including very distant ones, but it also pulls in a lot of false positives. Small segments are frequently identical by state rather than identical by descent, especially between people from the same broad continental population. Raising the threshold to 10 or 15 cM gives you fewer but higher-confidence matches.
There is no single “correct” setting. If you are trying to find relatively close relatives (third cousins or nearer), the 7 cM default works well. If you are investigating deep ancestry or trying to triangulate a specific segment, you might want to lower it cautiously. If you are swimming in thousands of matches from an endogamous population and need to find the ones most likely to be genealogically useful, raising the threshold filters out the noise.
The SNP density of the matching region also matters. A segment might technically span 8 cM, but if the region happens to be sparse in the markers that GEDmatch can read (because it depends on whatever chip your testing company used), the confidence in that match is lower. GEDmatch shows the number of SNPs in each segment on the One-to-One results page. Segments with very few SNPs relative to their cM size are less reliable, even if they clear the minimum threshold.
Privacy and Who Sees Your Data
GEDmatch is an open platform, and its privacy model differs from the locked-down approach of companies like 23andMe or AncestryDNA. When you upload your raw DNA file to GEDmatch, your genetic data becomes searchable by other users. Since 2019, GEDmatch has required users to actively opt in to law enforcement matching, meaning your kit is not visible to police searches unless you choose to make it so. But even if you opt out of law enforcement access, your kit is still visible to other GEDmatch users.
Security researchers have identified vulnerabilities in GEDmatch’s design. A study published in Briefings in Bioinformatics noted that researchers were able to demonstrate how an adversary could extract a large percentage of genetic markers from other users, including medically sensitive markers, by formatting data files and running standard queries through the platform’s open API.3Briefings in Bioinformatics. Digital DNA lifecycle security and privacy: an overview GEDmatch has since tightened some of its API access, but the underlying architecture remains more open than commercial testing services.
This openness is a trade-off. The same accessibility that makes GEDmatch powerful for genealogical research also means your DNA information is more exposed. If you are comfortable with that trade-off, the platform offers tools that commercial services do not. If privacy is a major concern, you should know what you are opting into before uploading.
How Forensic Genealogy Uses GEDmatch
GEDmatch became internationally known in 2018 when law enforcement used it to identify the Golden State Killer. The technique, now called investigative genetic genealogy, works by uploading DNA from a crime scene or unidentified remains and searching for partial matches in the database. Those matches are typically distant cousins of the person whose DNA was found. Genealogists then build family trees from the matches and work forward through public records to narrow down the identity.4Forensic Science International. Using genetic genealogy databases in missing persons cases and to develop suspect leads in violent crimes
The scale of this work has grown rapidly. As of recent published counts, at least 367 identifications of unidentified human remains have been publicly announced using investigative genetic genealogy methods, in addition to hundreds of criminal suspect identifications.5PubMed. Investigative genetic genealogy for human remains identification The process typically requires finding matches who share enough DNA to build a workable family tree, which means the technique is most successful when more people have uploaded their data to the searchable database.
For regular users, the forensic applications are worth understanding for two reasons. First, they explain why GEDmatch introduced the opt-in system for law enforcement searches. If you opt in, your DNA could be used as a stepping stone to identify someone else, even someone you have never met. You would not be accused of anything, but your genetic information becomes part of the investigative web. Second, the forensic use case illustrates just how much information is embedded in even distant DNA matches. A fourth cousin sharing 50 cM with a crime-scene sample can, in the hands of a skilled genealogist with access to public records, lead to a specific individual. That same power is available to you when you sit down with your own GEDmatch results and start building trees from your matches.
Tips for Getting the Most Out of Your Results
The single most productive thing you can do with GEDmatch is not to stare at the numbers but to contact your matches. A kit number with 200 shared cM is just a data point until you learn that the person behind it has a family tree going back to a village in County Cork or a plantation in Virginia. GEDmatch provides a messaging system for this purpose, though response rates vary widely. Matches who uploaded their data years ago and forgot about it may never reply.
When you do reach out, lead with specifics. Saying “we share 180 cM and our largest segment is on chromosome 7” gives the other person something to work with. If you have identified a possible shared surname or location through tree-building, mention it. Genetic genealogy is a collaborative endeavor, and the people who make the most progress are the ones who treat their match list as a network of potential research partners rather than a static readout.
If you have tested with a company like AncestryDNA or 23andMe, uploading your raw data to GEDmatch is free and expands your pool of potential matches to include people who tested with other companies or who uploaded data from older tests. GEDmatch accepts files from most major testing services, though the overlap of SNPs between different chips means some comparisons have lower resolution than others. A comparison between two AncestryDNA kits uses more shared markers than a comparison between an AncestryDNA kit and a 23andMe kit, which can slightly affect cM totals for the same pair of people. This is usually a difference of a few cM and rarely changes the interpretation, but it is worth knowing if you notice small discrepancies between platforms.