My sister and I recently did DNA tests through 23&Me.

We found out we have a surprise relative about 27 years younger than us (a generation younger). She has about 8.5% shared DNA or around 577 cM's shared with me. She has 10.2% shared DNA with my sister or about 694 cM's.

The crazy thing is that the dnageek website gives it an 81% chance that the relative is my 1st cousin once removed, while it tells my sister the same relative has a 67% chance of being a half niece. So we don't know if she is a 1st cousin once removed or are half niece and thus we don't know if her dad is our dad's kid or our uncle's kid and we can't get any more info.

Any thoughts out there on how to narrow this down from the shared dna data?

2 Answers 2


First of all, ignore the relationship predictions. While some relationships you can rule out or in based on what the segments look like, generally the predictions are guesses based on the cM total (and sometimes on segment size, for not-close family matches).

Check out the possibilities yourself: https://thegeneticgenealogist.com/wp-content/uploads/2017/08/Relationship_Chart_FINAL_August_2017.jpg

577 cM or 694 cM can be great-grandparent, half great uncle/aunt, half uncle/aunt, half first cousin, great great uncle/aunt, great uncle/aunt, first cousin, or first cousin once removed.

While you can rule out a couple of these just from context, you don't know yet about the rest. In particular, you don't know if it's a half first cousin or a first cousin once removed or even on the low end of the range for a regular first cousin. Then there are the aunt/uncle/niece possibilities.

What you need is more people to test or a clue on the paper trail. There are a few things you can do to narrow stuff down (for example, had you and your sister's cM totals with the other person been different in another direction, you could have ruled out a couple of the relationship possibilities.


The amounts of shared DNA you and your sister have with this relative both fall between the typical amounts for two particular groups of relationships. This is not surprising since the amounts are a result of a random process. The two groups involved here are the 12%-shared group and the 6%-shared group.

The 12%-shared group of relationships includes first cousin, great-niece, and half-niece.

The 6%-shared group includes first cousin once removed, half first cousin, great-great-niece, half-great-niece, and double-second-cousins.

You can explore the probabilities by plugging in shared cM amounts into the Shared cM Tool. Given your results, your best estimate would be from plugging in the average of your two cM amounts. Doing this shows a 60% likelihood of the relationship being in the 6%-shared group, a 39% likelihood of its being in the 12%-shared group, and a 1% probability of being in the 3%-shared group.

To resolve the relationship will require more knowledge of possible connections. This can come from whatever is known of your ancestry and your relative's ancestry, and can come from finding other people with test results showing matches with both you and your mysterious relative. To the extent you can figure out where these other matches fit in your ancestor tree and in your relatives tree, you can narrow down which of the relationships fit.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.