How DNA Relative Matching Actually Works

Upload your DNA to a testing service and, within days, a list of relatives appears - some you know, many you do not. It can feel like magic. It is not. Relative matching is a fairly mechanical process of comparing your DNA against everyone else’s, and understanding it reveals both how it works and what it quietly costs.

Comparing genotypes to find shared stretches

At its core, matching is a giant comparison. The company lines up your genotypes against every other person in its database and looks for shared stretches - regions where your DNA and theirs agree across a run of positions.

Because you carry two copies of each chromosome, these are usually found as half-identical regions: stretches where at least one of your two copies matches at least one of theirs, position after position. A long, unbroken half-identical region is the signature of DNA inherited from a common ancestor. The system scans for these runs across the whole genome, for every possible pairing.

Turning shared segments into a relationship

Finding shared stretches is only step one. The service then measures them - how many segments you share and how long each one is - and adds them up into a total, expressed in centimorgans (cM), a unit of genetic distance.

From that total and the pattern of segments, it estimates the relationship:

• Longer and more segments point to closer kin, because recent relatives inherited large, intact pieces from a shared ancestor.
• Shorter, fewer segments point to more distant relatives, since generations of shuffling have chopped the shared DNA into smaller fragments.

The prediction is a range, not a certainty, because different relationships can produce similar totals. Our explainer on centimorgans covers why one shared-cM figure usually maps to several possible relationships rather than a single answer, and why long identical-by-descent segments are the trustworthy ones.

Why bigger databases find more relatives

There is a simple reason large services surface more matches: matching only works against people who are already in the database. Your DNA can only be compared to someone if their DNA is there too. That is why the biggest databases return the longest match lists - not because they read your DNA better, but because they have more people to compare you against.

This is also the seed of the privacy problem.

The privacy cost built into matching

Relative matching is powerful precisely because so many people have placed their DNA into shared databases. And that is the catch: the system can expose people who never tested at all.

If enough of your relatives are in a database, your own genetic makeup becomes partly reconstructable from theirs - and the reverse is true too. When you add your DNA to a shared pool, you are also making inferences about your parents, siblings, children, and cousins possible, whether or not they ever consented. A single test can reach across a whole family.

This is not hypothetical. It is the same mechanism that lets investigators identify people through distant relatives’ DNA. The convenience of finding cousins and the risk of exposing relatives are two sides of the identical process. Our guide to genetic privacy for genealogy beginners goes deeper into how that exposure works and what you can do about it.

A more private way to explore

None of this means DNA exploration has to involve a shared database. If your goal is to examine your own data - or to compare two files you both control - you can do it without uploading anything. Genespiral lets you compare two DNA files privately in your browser, so the comparison happens on your device and neither file joins a pool that could expose anyone.

Relative matching is genuinely useful, but it is worth entering with open eyes: the same connectedness that finds your cousins can reveal family members who never chose to be found.

This article is educational and is not medical advice.