How Much DNA Do You Share with a Cousin? The Real Numbers Behind Your Match List

Did your DNA test just flag a total stranger as a likely relative, with 212 cM next to their name? So how much DNA do you share with a cousin? The closest ones, who share your grandparents, carry about 12.5% of your DNA, or roughly 866 centimorgans on average. Averages hide a lot, though, and that 212 cM still needs context.

Plenty of real cousins look like strangers on paper. A Brigham Young University study that simulated 100,000 pairs for each cousin type found that 98.5% of third cousins share at least one matching segment, yet only 32.7% of fifth cousins do. This guide covers a full cousin chart and a clear way to identify your match.

Cousin math mixes biology with family history, and it's easy to get lost halfway down a match list. That's where Nibble fits in, filling in the background a little at a time. Its Genealogy topic covers surnames and old records, while Biology lessons explain inheritance in the time it takes to finish a coffee.

Get the Nibble app and pick up family history basics one small lesson at a time.

Note: DNA test estimates are probabilities, and results vary between people and testing companies. Nibble is a microlearning and daily growth app, not a genetic testing, medical, or legal tool. For health questions tied to your DNA, talk to a doctor or a genetic counselor.

The short version comes first, and the details follow below.

• A 2nd cousin is expected to share about 3.1% of your DNA.
• Third and fourth cousins sometimes share no detectable DNA.
• Recombination shuffles DNA each generation and decides where your number lands.
• One cM total can fit several family relationships at once.
• Percentages differ between testing sites, so cM is the better number to compare.

Keep reading for the full chart and a step-by-step way to identify your match.

The figures below come from the Shared cM Project, Blaine Bettinger's crowdsourced dataset of confirmed relationships, and each row follows the same halving pattern as a standard cousin chart.

Each row lists one cousin type with its average share, its typical range, and the ancestors you have in common.

A 1st cousin match can land anywhere between 396 and 1,397 cM. That's a wide spread for one relationship, and a double shuffle causes it.

Your parent and your cousin's parent each got a different mix from the grandparents you share. You and your cousin then got a second mix. At 1,300 cM, a first cousin falls inside the half sibling range, too.

Nearly every second cousin shares a measurable amount of DNA with you, averaging 229 cM.

The link runs through your parents. Your mom or dad is a first cousin to theirs, and you both descend from the same pair of great-grandparents.

Third cousins connect through great-great-grandparents. Fourth cousins sit one generation further back, and both are expected to share under 1%.

Look at the low end of their ranges in the table. It's 0 cM, and that zero gets its own section below.

A half first cousin shares roughly half of what full first cousins do, because you share one grandparent instead of two. First cousins once removed land in a similar zone for another reason.

The label "removed" marks a generation gap, like your parent's first cousin. That gap is what separates removed cousins from second cousins your own age.

🧬 The app lets you chat with 10+ historical figures, including Marie Curie and Cleopatra. Ask one a question in the Nibble app.

Shared DNA is the set of matching DNA segments two people inherited from the same common ancestor. Everything else in genetic genealogy builds on that idea.

Each parent gives you half your DNA, so a parent / child pair always shares 50%. The expected share halves with every generation after that: 25% for a grandparent and grandchild, and half of that for first cousins.

So why doesn't every pair hit the average? Recombination. Before a parent's DNA reaches a child, the two copies of each chromosome trade pieces at random. Siblings end up with different mixes from the same mom, and that's why your brother might match a cousin you barely register with.

A centimorgan (cM) is a length unit for DNA segments. Add up the whole genome and you get about 7,000 cM, give or take a few hundred depending on the testing company. By that math, a 212 cM match is roughly 3% of your genome.

Plenty of things can pull your number away from the average. These five come up most.

1. Recombination luck: Two sets of first cousins in one family can end up at opposite ends of the range, purely by chance.
2. Company math: MyHeritage counts a fully identical segment only once. That's why its full sibling matches hover near 37.5% while other sites show about 50%, and why cousin percentages drift between 23andMe, AncestryDNA, and MyHeritage too.
3. Double first cousins: When two siblings marry two siblings, their kids share around 25%, the same as a half sibling pair.
4. Children of identical twins: Their parents share nearly all their DNA. Genetically, these cousins look closer to half siblings than to typical first cousins.
5. Endogamy: This means marrying within a small community for many generations, as in Ashkenazi Jewish and Acadian families. Cousins from these groups often share more DNA than the chart predicts, which can make family relationships look closer than they are.

Yes, a real third cousin can share no detectable DNA with you, and the odds of a blank grow the further back your shared ancestors sit.

Every test has a cutoff. 23andMe only counts a match once a shared segment reaches at least 7 cM, so shorter pieces get tossed even when a real ancestor handed them down.

That cutoff costs you relatives. The company estimates it catches about 90% of 3rd cousin pairs and about 45% of 4th cousin pairs, so plenty of real cousins never appear in your DNA matches.

Genealogists have a name for this: negative evidence. If one documented 4th cousin doesn't show up, you can shrug it off.

Ten no-shows from the same couple are harder to wave away. At that point, something in the paper trail probably broke, like an adoption nobody mentioned or two records that got mixed up.

Your ancestors double with each generation, so 10 generations back you have 1,024 of them. As a result, thousands of distant cousins share a common ancestor with you but no detectable DNA.

Population models suggest that a few thousand years back, shared ancestry links every person alive today. Some family lines spread far wider than the rest, like the Mongol lineage often tied to Genghis Khan.

🧬 Audio episodes run close to 10 minutes, about the length of a short walk. Press play on the Nibble app.

Totals in cM overlap between relationships, so one number can fit several relatives.

Parents, siblings, grandparents, and aunts or uncles share far more DNA than cousins, yet their ranges still overlap with each other.

Closer relatives overlap just as often. A match at 1,750 cM could mean a half sibling, a grandparent / grandchild, or an aunt/uncle.

Say Maria, 64, shows up with 871 cM, and you're 31. That total fits a 1st cousin, a great-grandparent, a great-uncle / aunt, a half-aunt, or a half-niece.

Her age rules out great-grandparent and half-niece. A 33-year gap also makes first cousin less likely, so a great-aunt or half-aunt moves to the top.

Age difference is the quickest clue. Shared matches come next, and your family tree settles most of what's left. Free online cM calculators also show the odds of each relationship for your exact total, which saves guesswork in family history research.

Free tools and a little patience go a long way here. Work through these steps in order.

1. Write down the cM total and the number of segments. Percentages vary between sites, so skip them.
2. Enter the cM total in DNA Painter's Shared cM Tool to see which relationships fit.
3. Open your shared matches and note which branch they cluster on.
4. Compare family trees for surnames and places that repeat.
5. Build your family tree back to your great-great-grandparents. That covers every common ancestor you share with third cousins, and census and vital records, the genealogy basics, can fill in many of those names.
6. Send the match a brief message naming the ancestor couple you likely share.
7. Ask an older relative, like a parent, aunt/uncle, or grandparent, to test too, which reaches one generation further back.

Database size matters as well. AncestryDNA and 23andMe have the largest pools, so taking a DNA test where your relatives tested improves your odds.

Every major testing company shows a shared matches list. It's often the fastest way to sort DNA relatives by family side.

Some people carry two different sets of DNA

Nibble digs into more DNA quirks like this

Explore Biology

So how much DNA do you share with a cousin? It depends on how many generations sit between you, plus a fair bit of recombination luck. Your cM total is the first clue, and your tree does the rest. Nibble gives you an easy way to keep building on it.

The app asks what you're curious about and builds a "For You" feed around it. Genealogy can sit next to Biology or History, with lessons of 1.5–15 minutes. Match the Pairs quizzes you on what you read, while a daily streak nudges you back on busy weeks.

Spare phone minutes add up, and these leave you with something to share at the next family dinner. Bea, a nutritionist who made Nibble part of her evening routine, sums it up: "It's surprising how much you can learn in 10 minutes."

Pick your first lesson in the Nibble app tonight, and give those family tree questions a daily home.