I recently found a photo of my maternal grandparent’s wedding party. I’ll refer to my maternal grandparents as GM (grandmother) and GF (grandfather). I was surprised to see that one of the people in the wedding photo (who is not the person GF that I’ve always believed was my grandfather) looks very much like me. I’ll refer to this other person as MP. Also, my mother is the oldest of three sisters, and while her two younger sisters (S1, S2) look alike, my mom doesn’t share many of their traits. My obvious thought - is it possible that my mother is the offspring of GM and MP, while my mom’s two sisters are the offspring of GM and GF?

I recently took a 23andMe test. When I got the results, there was only one person listed as a relative that I recognized - the son of my mom’s sister (S1), who I’ll call 1C. This is the person that I’ve always known to be my 1st cousin. However, in the 23andMe results, he’s listed as being my “1st cousin once removed” saying that we share 9.23% DNA, 26 segments. (Specifically it says that 1C could be the child of my 1st cousin, or parent of my 2nd cousin.) There a lot of other relatives listed, but no one else listed as being a closer relative (the next closest was a 2nd cousin, 2.25% DNA, 7 segments), and I didn’t recognize any of the names in the list (other than 1C).

So my question - does the fact that the 23andMe lists us as being "1st cousin once removed” prove that we’re not actually 1st cousins, or do results like this have a degree of uncertainty such that this can’t be considered proof?

A second question - I don’t know who MP is, but it’s possible that he could be the brother of my GF. How would that affect the genetic determination of how 1C is related to me (as opposed to if MP were a complete stranger and not related to GF)?

I know that this could be resolved by having my mom and S1 (and S2) take 23andMe tests to see if they are all full sisters, but as far as I know none of this is known in the family, and I don’t want to bring it up. I’m just interested for my own curiosity. Thanks for any help.

I don’t know if this helps or is relevant, but additional information from 23andMe:

Maternal Haplogroup: Me = T2b   , 1C = T2b
Paternal Haplogroup: Me = J-L24 , 1C = I-Z58
  • Hi, welcome to G&FH.SE! Since "OP" is used heavily across the Internet to refer to the "original poster" (in this case, you, the person asking a question), I have taken the liberty of editing your question to replace OP with MP (for "mystery person"). I've also added some white space to make your question easier to read.
    – Jan Murphy
    Commented Feb 16 at 1:29
  • Thanks Jan! Appreciate the help and the welcome.
    – MrAnon1900
    Commented Feb 17 at 3:25
  • While you are waiting for answers, take a look at other DNA questions on the site to see strategies for analysis in the answers there, standard references on DNA like the ISOGG Wiki and recommended books, and other information that might help you.
    – Jan Murphy
    Commented Feb 17 at 4:26
  • If you know the cM amounts that you share with 1C and the others you mention, you can use DNA Painter's shared cM tool <dnapainter.com/tools/sharedcmv4> to figure out possible relationships.
    – user5836
    Commented Feb 23 at 15:19

1 Answer 1


Any "half" relationship has a corresponding "once removed" relationship with a similar match percentage (for example, "half sibling" corresponds to "niece/nephew"). In the interests of not making accusations of marital infidelity, genetic-genealogy sites will usually omit the "half" option in favor of the "once removed" option.

The data you've got can't rule out any of the possibilities. A match of 9.23% (about 687 centimorgans) is unusually high for both half first cousin and first cousin once removed. Only about 5% of such relationships will have numbers that high or higher. However, it's also unusually low for a full first cousin (about 13% of cases are that low or lower).

In order to remove the uncertainty through genetic testing alone, you'll need to test more people. "Mother" and either "S1" or "S2" would be conclusive -- there's nearly no overlap between the expected match percentages for "full sibling" and "half sibling". However, testing additional descendants of "S1" or "S2" would also provide evidence one way or the other.

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.