3rd Cousin Explained: The DNA, the Matches, and the Detective Work
A 3rd cousin is someone who shares a set of great-great-grandparents with you, with each of you sitting exactly four generations below that couple. On paper it's a simple definition. In practice, the third cousin is the most interesting and most frustrating relationship in all of genetic genealogy, because it sits right at the edge of what a DNA test can reliably see.
That edge is the whole story. Third cousins are numerous enough to fill your match list, close enough that identifying them can crack a brick wall, and distant enough that roughly one in ten of them will not show up in your results at all. If you've ever stared at a page of "3rd–4th cousin" estimates wondering what to do with them, this is the guide for that.
The Definition and the Counting Rule
To identify any cousin relationship, count how many generations each person is from their most recent common ancestor. The cousin number is the smaller count minus one, and the removal number is the difference between the two counts.
For third cousins, both people are four generations from the shared ancestor. Four minus one is three, so third cousins. The difference between four and four is zero, so nobody is removed.
Walk it up the ladder and it's easy to keep straight:
- Share grandparents (2 generations each) → first cousins
- Share great-grandparents (3 generations each) → second cousins
- Share great-great-grandparents (4 generations each) → third cousins
There's an even faster shortcut that works every time: your parents' second cousins' children are your third cousins. Each generation moves the cousin number up by one, provided both lines descend at the same rate. If the definitional side is what you're after, the second cousin guide covers the family-relationship angle in more depth.
If you'd rather skip the counting, the free cousin calculator on cousinchart.com will map any two positions on a family tree to the right label, removals included.
How Many 3rd Cousins Do You Have?
Most people have hundreds of third cousins, with common estimates landing anywhere from 175 to 500 or more, and people from historically large families easily passing 1,500.
The spread in those figures isn't sloppiness. It reflects how completely this depends on family size in each individual branch. To count your third cousins, you'd need every descendant of all sixteen of your great-great-grandparents, minus the closer relatives. Sixteen founding couples, each with however many children they had, and every one of those children's lines multiplying for three more generations.
Rough figures worth keeping in mind, all of them soft estimates:
- Roughly 20 to 30 first cousins
- Roughly 150 to 200 second cousins
- 1,500 or more third cousins for people from large families, and a few hundred for people from small ones
If your ancestors farmed and had eight children per couple, your third cousin count runs into the thousands. If your family tree passed through a famine, a war, or a couple of generations of only children, it might be under a hundred. There is no universal number, and any article that gives you one without hedging is guessing.
The practical consequence is huge for DNA testing: you have so many third cousins that a good number of them have almost certainly tested. You just don't know their names.
The 3rd Cousin DNA Numbers
Third cousins share about 0.78% of their DNA on average, roughly 73 centimorgans, but the observed range runs from 0 all the way to 234 cM. That zero is not a typo, and it's the single most important fact on this page.
| Relationship | Shared ancestor | Avg. shared DNA | Avg. shared cM | Observed range (cM) | Share any detectable DNA |
|---|---|---|---|---|---|
| 2nd cousin | Great-grandparents | ~3.125% | ~229 | 41–592 | ~100% |
| 3rd cousin | Great-great-grandparents | ~0.78% | ~73 | 0–234 | ~90% |
| 4th cousin | 3x-great-grandparents | ~0.195% | ~35 | 0–139 | ~50% |
| 5th cousin | 4x-great-grandparents | ~0.049% | ~10–25 | 0–~117 | ~10–15% |
The observed ranges come from the Shared cM Project, Blaine Bettinger's crowdsourced dataset assembled from tens of thousands of documented relationships. It's the honest citation for this kind of number, because it reports measured results rather than theoretical predictions.
For scale, the full measurable genome on consumer tests runs to roughly 6,800 cM, with about 3,400 to 3,500 cM coming from each parent. A 73 cM match is under 1% of that.
Read the last column of that table carefully, because it's what separates people who use their match list well from people who get frustrated by it.
About 10% of Real Third Cousins Share No DNA at All
Roughly 90% of true third cousins share detectable DNA on an autosomal test, which means about 1 in 10 genuine third cousins will show absolutely no match. They are related, documented, and invisible to the test.
This surprises people, so here's why it happens.
You inherit exactly 50% of your DNA from each parent, and that's the only step in the whole chain that's fixed. Everything after that is random. When your parent passed DNA down to you, recombination shuffled and cut their chromosomes, and the pieces that survived were a coin flip. Your third cousin's parent did the same shuffle independently.
Trace that back four generations. Each of your sixteen great-great-grandparents contributed about 6.25% of your DNA on average, but the actual figure for any one of them might be 10% or might be 2%. Now the same lottery ran on your third cousin's side. The chance that you both happened to keep an overlapping segment from the same ancestor, long enough for the testing algorithm to call it a match, is about 90%. Good odds, but not a guarantee.
Three things follow directly, and they matter:
- A missing match is not evidence against a paper trail. If your documented third cousin doesn't appear in your results, the most likely explanation is ordinary chance, not a hidden misattributed parentage.
- A DNA match is much stronger evidence than a non-match. Presence proves relatedness. Absence proves very little at this distance.
- Test more relatives. If your sibling, parent, aunt, or uncle tests, they inherited a different shuffle. A third cousin invisible to you may show up clearly for them, and each additional close relative you test meaningfully widens the net.
At fourth cousins this problem becomes dominant. Only about half of true fourth cousins share detectable DNA, so a fourth cousin match list is a coin flip on every relationship. Third cousins are the last level where you can plan around the results.
Why Your Match List Is Full of "3rd–4th Cousin"
DNA testing companies label enormous numbers of matches as "3rd–4th cousin" because that estimate covers a very wide cM band that dozens of different real relationships fall into.
The prediction is generated from one number: total shared centimorgans. The company sees, say, 60 cM and reports the statistically likeliest bracket. But 60 cM is genuinely compatible with a third cousin, a third cousin once removed, a fourth cousin, a half second cousin once removed, and several other configurations. The company isn't being vague to annoy you. The data honestly does not distinguish them.
This is why the 3rd–4th band is so crowded:
- You have hundreds or thousands of third cousins and tens of thousands of fourth cousins, so there are simply more people at this distance than at any closer one.
- Enough of them share detectable DNA to appear, but not enough DNA to be pinned down.
- The cM ranges for these relationships overlap almost completely.
The useful mental reframe: treat the company's label as a rough sorting bin, not an answer. The number of centimorgans, the number of shared segments, and above all the shared matches are where the real information is.
How to Actually Identify a Third Cousin Match
The reliable way to identify a third cousin match is to cluster them with matches you already know, then work down from the great-great-grandparent couple that cluster points to. The tree work does the identifying; the DNA only tells you where to look.
Here's a workflow that works:
1. Build your tree back to all sixteen great-great-grandparents. This is non-negotiable. A third cousin connects through one of those sixteen people, so if your tree stops at grandparents you have no place to attach anyone. Most stalled DNA research is really stalled paper research.
2. Use shared matches first, always. Open the mystery match and look at who else you both match. If they share matches with your known paternal grandmother's line, you've cut the problem from sixteen ancestral couples to two. This single feature outperforms every other tool on these sites.
3. Cluster the whole list, not one match at a time. Grouping matches into clusters by shared-match overlap turns a chaotic list into four to eight family groupings, each pointing at one branch. Many sites offer colored dot labels or automated clustering for exactly this.
4. Read their tree sideways, not top-down. Your third cousin's surname often has nothing to do with yours, because the connection can run through three consecutive maternal lines. Look at the surnames in their great-grandparent generation and compare against yours.
5. Check locations and dates before names. Two families in the same small county in the same decade is a strong lead even when no names match. Third cousin connections often surface first as a place, not a person.
6. Watch for endogamy. If your ancestry comes from a historically intermarried population, cM totals run inflated and a "third cousin" match may be a more distant relative connected through several lines at once. Adjust your expectations downward on cM in those trees.
7. Message with something specific. "Hi, we share 71 cM and my great-great-grandparents were Samuel and Lena Hoffman of Wayne County, Ohio, born about 1855. Do those names appear in your tree?" That gets replies. "Hi, how are we related?" mostly doesn't.
Third Cousin Once Removed and Twice Removed
A third cousin once removed is either your third cousin's child or your parent's third cousin, because "removed" describes a generational offset rather than extra distance sideways.
Apply the counting rule. One person is four generations from the shared great-great-grandparents, the other is five. The smaller count is four, so four minus one gives third cousin. The difference is one, so once removed.
| Relationship | Generations from MRCA | Avg. shared DNA | Avg. shared cM | Observed range (cM) |
|---|---|---|---|---|
| 3rd cousin once removed | 4 and 5 | ~0.39% | ~48 | 0–192 |
| 3rd cousin twice removed | 4 and 6 | ~0.195% | ~22 | 0–~130 |
Two directions, same label:
- Downward: your third cousin's children are your third cousins once removed. Their grandchildren are your third cousins twice removed.
- Upward: your parent's third cousins are your third cousins once removed. Your grandparent's third cousins are your third cousins twice removed.
For DNA work, removals matter mostly because they widen the field of candidates for any given cM number. A 48 cM match is a perfectly good third cousin once removed, and also a perfectly good fourth cousin, and also a plausible third cousin who got unlucky in the shuffle. This is why you never identify a distant match on centimorgans alone.
Are Third Cousin Matches Worth the Effort?
Third cousin matches are worth pursuing when they cluster around a specific brick wall, and generally not worth pursuing one at a time out of curiosity.
The honest cost-benefit:
Worth your time when:
- The match clusters with a line where your research is stuck
- They have a public tree going back four or more generations
- Their tree shows a surname or location you've been chasing
- You're trying to confirm a specific ancestral couple through multiple independent descendants
Probably not worth your time when:
- The match has no tree, no response, and no shared matches you recognize
- You already have the line well documented on paper
- You're working through the list alphabetically instead of by cluster
Confirming an ancestral couple through several independent third cousin descendants is genuinely powerful evidence, far stronger than any single match. That's the payoff worth working toward.
Frequently Asked Questions
How much DNA do 3rd cousins share?
About 0.78% on average, roughly 73 centimorgans. The Shared cM Project records observed values from 0 to 234 cM, which is a very wide spread. That variability is normal, because DNA inheritance is randomized at every generation past parent to child.
Can third cousins share no DNA?
Yes, and it happens roughly 10% of the time. About 90% of true third cousins share detectable DNA on an autosomal test, so around one in ten genuine third cousin pairs will show no match at all. A missing match at this distance says nothing negative about your paper trail.
Is a third cousin close family?
Genetically, no. You share less than 1% of your DNA on average, and legally and socially third cousins are treated as distant relatives everywhere. That said, plenty of families stay in close contact with third cousins through reunions or shared hometowns, so closeness in practice has more to do with contact than with centimorgans.
What is the difference between a third cousin and a second cousin once removed?
A third cousin is four generations from your shared great-great-grandparents, and so are you. A second cousin once removed is three generations from the shared ancestor while you are four, or the reverse. The second cousin once removed shares roughly double the DNA on average, about 122 cM against 73 cM.
Why does my DNA test say "3rd–4th cousin" for so many people?
Because that label covers a broad centimorgan band that many different real relationships fall into, and because you simply have more relatives at that distance than at any closer one. The estimate is a sorting bin, not a conclusion. Use shared matches and tree research to narrow it down.
Can third cousins marry?
Yes. Third cousins can marry legally throughout the United States and in essentially every country that permits cousin marriage at all, since every restriction that exists applies to first cousins or closer. Third cousins share under 1% of their DNA, so the genetic considerations are no different from those for an unrelated couple.