Building a Family Tree from DNA: From Matches to Documented Ancestors

A DNA match list and a family tree do not, by themselves, build each other. The matches are facts about shared DNA. The tree is a documented genealogical structure. Connecting one to the other is the actual craft of genetic genealogy. Done well, it turns a confused list of cousins into verified branches that go back many generations. Done badly, it produces confident-looking trees that are wrong.

What a tree built from DNA actually proves

A DNA match between two people proves they share a recent common ancestor. It does not name that ancestor. The math gives you a relationship range, not a relationship. The tree side of the work, documents, census records, birth and marriage certificates, is what turns a 230 cM match into “we are second cousins through our great-grandparents Henry and Mary.” You need both halves.

This is why every serious genetic genealogist treats DNA evidence and document evidence as complementary. A documented tree without DNA can be wrong in subtle ways the documents will not catch, such as a non-paternity event or an adoption never recorded. A DNA-only tree without documents is a hypothesis, not a conclusion.

Start with one anchor and work outward

The standard workflow starts with the closest matches who have public trees. Find a match in the second cousin range or closer, look at their tree, and identify the most recent ancestral couple your DNA implies you share. The relationship range from our shared DNA and relationship chart tells you which generation that couple is likely to be in.

From that anchor, work outward. Other matches who descend from the same ancestral couple should appear in your match list at cM amounts consistent with their distance from that couple. When they do, you have triangulated a branch, and the documented descent from that couple becomes part of your verified tree.

Match clustering as a tool

Match lists are easier to read when you cluster them. Most close matches share matches with each other in a pattern that reveals which side of your family they connect through. A cluster that all share each other and share you, but does not share another cluster, almost certainly represents one side of your family.

AncestryDNA’s SideView, MyHeritage’s AutoClusters, and various third-party tools at GEDmatch all automate clustering to some degree. The output is not a family tree, but it is a useful first cut at which matches connect through which ancestors.

ThruLines and Theory of Family Relativity

AncestryDNA and MyHeritage both offer features that propose common ancestors between you and a match, based on their respective sites’ public trees.

AncestryDNA’s ThruLines uses the trees attached to your match cousins’ accounts and your own tree, walks back through them, and proposes ancestors you appear to share. ThruLines suggestions are not proven; they are hypotheses that depend on the accuracy of the trees feeding them. Treat each ThruLines suggestion as a candidate to verify with primary documents, not a fact.

MyHeritage’s Theory of Family Relativity does the same kind of work using its trees, its Smart Matches, and its records. The same caution applies. Both tools are extraordinarily useful when they work, and both are wrong often enough that you should never trust them blindly.

Where to host the tree itself

A documented family tree lives in a tree service. The three most-used in the consumer space are:

• Ancestry’s tree service is the largest by far and integrates tightly with AncestryDNA matches.
• MyHeritage’s tree service is the strongest in Europe and integrates with MyHeritage DNA.
• FamilySearch is free, run by the LDS Church, and has an enormous shared global tree that anyone can edit.

You can keep your tree in more than one of these. Tools like GEDCOM exports let you move tree data between them. For a search-focused tree, host on Ancestry because that is where the largest match base lives.

How long does it take

A first-time tree built from DNA matches, with no prior research, usually takes weeks to get the first several generations solid and months to push back to the 1800s or earlier. Tree work compounds: once you have a few generations documented and a clustering of matches that confirms them, extending another generation back becomes much faster.

Three habits make the difference. Record sources for every claim, even when you are confident. Re-verify ThruLines and Theory of Family Relativity suggestions before you write them into your tree. And when DNA contradicts documents, follow the DNA. The documents may have a clerical error, a hidden adoption, or a non-paternity event. The DNA is harder to fool.

For unexpected results that emerge in this process, see unexpected DNA results: what to do. For the mechanics of cousin matching that underpins all of this, see how DNA matching works and understanding centimorgans. For the pillar overview, see the ancestry and genealogy guide.