Full versus Half Sibling DNA Matches

Recently I was asked whether there is a way to tell a full sibling from a half sibling. Sometimes the total centimorgans are low for one but high for the other. You expect to share about 2550 cM with a full sibling and 1700 with a half sibling; so which is it if you share 2100 cM?

The answer is that full siblings will share many fully identical regions (FIRs), over about 25% of their chromosomes. Half siblings will have no FIRs of any significant size. The exception being some small ones if their parents are from the same population group but still far far fewer than a full sibling.

Here is a comparison of my first eight chromosomes with my brother made at GEDmatch. The green bars are where we are fully identical, in other words we got the same segment of DNA from each of our parents there.

GEDmatch comparison of me and my brother

GEDmatch comparison of me and my brother (full siblings)

You can click on the image to see the entire display at GEDmatch that this was made from.

The old 23andme used to show these FIRs and HIRs in the “Family Traits” display when you selected Genome-Wide Comparison. You can see my brother and mine on this page: http://blog.kittycooper.com/about-me/our-dna-test-results/autosomal-dna-picture/

I have not been able to find this tool in the new 23andme experience, so I guess it is gone. If anyone finds it let me know. There is a new way it shows traits under Share and Compare and many new traits are listed. For example, most of my family has a widow’s peak. [UPDATE 5 Apr 2017: FIRs are now back, they are shown when you compare DNA relatives in the chromosome browser.]

So how to get the FIRs from GEDmatch? In the one-to-one comparison be sure to you click “Yes” next to the question “Show graphic bar for each Chromosome?” You will get a display of every chromosome (not in scale to each other) and the numbers that go with the matches as well as the blue bars for each half matching region (HIR) and the green bars for the FIRs. [UPDATE: you can also click next to “Graphic Only” to get just the picture]

A word of warning, double first cousins, where two brothers married two sisters, will also have many FIRs (about 12.5%) but their centimorgan count will look like half siblings.

Use the article at ISOGG on autosomal statistics as your reference guide for the expected centimorgans for various relatives: http://www.isogg.org/wiki/Autosomal_DNA_statistics


UPDATE 3 jan 2017: Now that you can look at just the image at GEDmatch I thought it might be useful to look at a pair of known half siblings (click for a larger image)… no large green bars.


Half Siblings at GEDmatch (click for full image)

151 thoughts on “Full versus Half Sibling DNA Matches

  1. Half siblings have the same father or the same mother. Therefore they most likely will share some fully identical regions (FIRs) from their common parent. How do you explain your comment that “Half siblings will have no FIRs of any significant size.”?

    • Carl,
      Perhaps I should have defined my terms better. Fully identical means a segment with a match on both the paternal and maternal chromosomes.

      You have 23 pairs of chromosomes, 23 from each parent. So a half sibling will only share DNA with you from one parent, thus half indenticall in those places. Full siblings will have sections where they got the same DNA as each other from each parent thus fully identical sections.

      See if this explanation of FIRs at ISOGG helps

      • I just got my test results back and I’m having a little trouble understanding them so does 45% mean that me and the alleged brother are half siblings or not

        • Did you get a reply? I have a similar situation 40% DNA match with a half sibling. Would you share what you discovered about this?

          • Bonnie – same answer, likely a full sib but look at the image for FIRs. A half sib would be only 25% but if the the other parents are related it could be higher.

        • Where did you test – better to look at the total cMs and whether there are any FIRs.
          45% at 23andme is a full sib, a half sib would share about 25%

    • Kitty, I just completed DNA testing for half sibling sharing a father (Trying to determine my bio dad is). I used TestMeDNA, they use LabCorp and the results came out with Inconclusive and show a percentage, a male with 32.9 percent and a female 32.4%. Their dads are brothers, yet, my match is almost identical to both. Not sure I understand that, other than the male with both XY conclusion may be stronger. My belief is that the female is my half sister and her father being the same as mine. They said anything under 91% they give an Iconclusive as the answer. I saw no cMs or anything mentioned in your blog to compare. My test results show things like HB1. HS1 D3S1358. D7S820 vWA FGA D8S1179 DS21S11 D18S51 D5S818. D13S317. D16S539 Second set THO1. TPOX. CSF1PO. AMEL. Penta D. F13A01. F13B LPL. Penta E. FESFPS. and last box. HB1. HS1 PI. It also says the sibling indeedx for Half siblings is 0.49-1. This value is inconclusive Are you able to help me figure this out? Did I take the wrong DNA test? Thank you in advance

  2. Interesting, didn’t realize double 1st cousins could produce that.

    The first half-brother I got tested turns out we share 1602.5cM. The other half-brother I tested we share 1981.8cM.

    When my mother’s new found sister (actually my mother is the new found sister) asked to get DNA tested I was happy. But her reason was to see if she was a half-sister rather than a full-sister as been believed by the entire family. My aunt felt that her father was not her biological father. I thought that would be easy to determine.

    Initially I saw the results at FTDNA where they came up predicted as full or half sibling sharing 2182.92cM, about the same as what you mentioned in your example. I thought since my aunt was questioning her father, it would be easy to tell by their X. It was actually confusing!

    I immediately transferred over to GEDmatch (the next day since there were issues at the time with GEDmatch) and saw that they shared 2084.8cM, not much of a difference. But still the X chromosome was confusing to me. I expected either a full 196cM or so on the X or significantly less.

    Luckily after doing a comparison on the autosomes it was obvious, although I expected several FIR but in small amounts, turns out they really are half-sisters and my aunt was right.

    You’ll see what I mean by confusing with the X in the diagram.


  3. Thank you Kalani for sharing the GEDmatch image for half siblings in an endogamous population!

    The small FIRs are as expected since the farhers are related, albeit distantly multiple times.

    You were right to think the X would tell you. people sharing the same father get the same unrecombined X from him so a solid blue bar on their X as per this article http://blog.kittycooper.com/2014/03/how-can-the-x-chromosome-help-with-maternal-versus-paternal/

    Your sample has some tiny breaks on the X (chr 23). Possibly since the mother’s parents would have been distantly related, she had many FIRs on her X leading to what looks like less recombination

    • Thanks Kitty for that. I couldn’t understand why the X would look like that. At FTDNA, it looked differently of course, looked more like slices were taken out.


      I never gave it any thought that my mother’s maternal grandparents may have been distantly related multiple times which is also a possibility. Although my mother still has two more half-brothers alive who could test, both have different fathers and probably from my aunt too, it would be interesting to see how much X they all share with each other.

      • Kalani
        It would be very interesting. Please let me know. The X fascinates me. Will I see you at Jamboree?
        One mistake I see people make is assumin a small X match is on the same line as the large autosomal match. This really does not have to be so, especially in even mildly endogamous populations!

  4. Interesting blog.

    I just read your message on the Newbie group and thought I might learn something.

    We are trying to determine if two women who have the same mother, might also have 2 brothers as fathers. It’s so confusing. I was hoping that xDNA would help determine this. I’m posting a link to the GEDMATCH data that I produced for them. I could really use some help figuring this out.



    • I think CeCe answered you on the Newbies list. The lack of green bars makes it unlikely. Look also at their autosomal DNA as well. Again there will be green bars if the men are brothers.

  5. Thanks,

    It just dawned on me that it would be difficult to see if two brothers fathered half sisters using xDNA since they got their xDNA from recombined xDNA from their mother. I assume that it recombines each time a new child is conceived. So we can only prove the 1/2 sibling relationship based on 28% shared autosomal.

    I’m learning more every day.

  6. Thanks for this. My half brother has tested and is waiting for his results (from AncestryDNA). I look forward to explaining to him why I’d like him to export his results to GEDMatch (or FTDNA) so I can see the results on a chromosome browser. Since I have no first cousins on my dad’s side, this will be a great help going forward in my DNA learning.

  7. Hello Kitty:

    My half sister is showing as sharing 1484 cms. Ancestry is showing us as first cousins as well as Gedmatch, but there is no way that she is my first cousin because my dad dated her mom.

    • I have a similar issue sort of. When comparing my DNA in GED match, someone came up with 1293 as a score saying first cousins. The issue with that is that all of my parents siblings lived in CT and were under 18 and my mother and father were split at the time but lived near where the person was born and raised. Further we added each other on FB and he is a dead ringer for my father and his brothers. Wondering if the amounts are lower for opposite sex half siblings because honestly, there are no mystery children that would be siblings to my parents that would have been old enough to parent a child.

      • That is just below the bottom of the range for a half sibling but not impossible if you are tested at different companies. Can you get more paternal side relatives to test to help narrow this down? Get in touch with your new “cousin” and offer to help. Adoptees are told to be circumspect with mentioning that because many will turn away at that news. Sounds like you are not one to do so! Congrats

  8. Troy
    That total is a little bit low for half siblings but certainly possible and very high for first cousins. Ancestry and GEDmatch have no way of knowing which it is, they are guessing from the match size. Look at the statistics here to see all the possibilities:
    There is no easy way to prove a half sister to a half brother who share a Dad. However siblings will usually have larger segments of matches (50/60/70/80 cM) than first cousins. So look at those. What are the largest two segments in the match?
    If there are other possible relationships, see if you can disprove those in the DNA.

  9. Thanks for your response. The largest segments are 165, 136, 131, and 108. I also have a few that are in the 70s and 80s. My dad told me that it was a possibility that she was his daughter. I feel 95% sure that she is but I wanted to be sure.

    Thanks again!

  10. Thanks, but I guess I need to know how you can tell from the length of the centimorgan? Sorry for asking so many questions.

    • Only very close family will share such large segments as you report. A cousin will not have many segments greater than 40cM. Only a sibling would share that many big pieces with you.

  11. If half siblings txt at around 1700 cm to 2000 .. thats a little daunting . my mothers brother and i share 2000 cm. a little strange that we had such a high number and my half sister shares 2267. I guess its confusing.

    • Linda
      Autosomal DNA has very few absolutes… is it at all possible that you are from an endogamous population group? That would explain these higher numbers. Also if your parents are distantly related.

  12. hello kitty i am doing a DNA testing for court for half sibling. My dad is deceased and the girl he raised which is not his daughter and me which i am his son are orderd by the court to do the half sibling test . question is how accurate is dna testing and will it show that we dont have the same father in the result. thank you for your response.

    • Muellew –
      Sorry for the late response but the reply I wrote on my phone today got lost somehow.
      If you are testing to see if you are closely related or not at all, then an autosomal DNA test, like Ancestry.com DNA or Family Tree DNA will answer your question. But you may need to upload the results to GEDmatch.com or DNA.land to see the details and confirm.
      Half siblings look very much like uncle/niece or grandparent/grandchild, so if those relationships are possible you may need further testing, either of cousins or a more traditional sibling test using STRs.
      However this adoptee, whose opinion I totally trust, suggests that those STR tests are not useful here:

    • Muellew – the courts will set up your ‘court ordered’ testing. But without a parent to compare to, I don’t see how they will be able to do this, as with my own paternal testing, each child and each parent had to submit DNA (done by buccal swab test)
      For more information on the legality of this, see the link: https://dnatesting.com/paternity-testing-without-the-father/

      But to add in my own .02 – both have done via Ancestry & FTDNA – my sons whom have different fathers, do show they are closely related. Both companies ‘suggest’ they are half siblings/cousins. And of course, because I also have my testing done at both companies, they both show me as their parent. So conclusive on our end – they cannot say I am ‘NOT’ their mama (as a pun towards Maury)

  13. thank you so kitty for your resopnse. Will the half sibling DNA testing show if we are related or not right Kittyeee?
    thanks again.

    • Sorry muellew, my expertise is for personal genome tests. I have no experience using those other tests, they were suggested as an alternate by another genetic genealogist

  14. I was contacted by an adopted girl who shares DNA with me. At first we thought she was a cousin according to ancestry.com, but according to timeline of her birth, we are thinking she may be the daughter of my half sibling I didn’t know about. We share 1106 cm and largest is 152. Is this possible that she is my niece rather than cousin

    • Tami, Yes either relationship is possible based on those cMs. Do you have other family members who are tested? Can you get more of them to test? can you ask, delicately, around in your family? If you have uploaded both kits to Gedmatch I would take a quick look for you, send me the kit numbers via my contact form

  15. Hi kitty…I just downloaded mine and my suspected to be half siblings/cousin’s dna in gedmatch. It’s was 1,377 cM with the largest segment being 128.1 cM. Would you say we were half siblings or first cousin.

    • Shannie –
      That is not enough information. It could be either or an aunt/niece relationship. Is the second largest segment also over 100? Are there any full chromosomes matching? How much X?
      Try the calculator at DNAadoption.com and I emailed you so that you can send me the kit numbers.
      The only way to be sure is to test more relatives of the non adoptee. For example, test maternal and paternal first cousins. If both match then aunt/niece.

  16. X and O’s to show just how confusing it can be but if you take the time
    to figure it out it can add a lot to your research.

    Some real life X DNA matches between half siblings, Full siblings
    half niece, full niece, half 1st cousin.

    K, D, J, G, and S’s deceased mother all have same mother
    G has different father
    K has different father than G and different than D,J, and S’s mother
    D, J, and S’s deceased mother all have same father
    D, J, and G are male

    S=Full niece to D,J, through deceased full sister, half niece to G and K

    Dd=half nephew to D and J and S’s mother (1/2 1st cousin to S) through
    D and J’s father and S’s grandfather. Dd’s mother has same father as D, J, S’s deceased mother but different mother.

    D-S X 128,500,991 133,915,684 5.1 541
    J-S X 128,500,991 133,968,536 5.2 246
    G-S X 128,500,991 133,968,536 5.2 247
    K-S X 127,872,771 134,326,312 6.2 652

    X 2,710,157 12,647,010 26.4
    X 13,142,754 44,399,427 49.8
    X 123,194,758 146,604,866 38.5
    X 10,024,609 44,399,427 54.5
    X 123,194,758 146,604,866 38.5
    X 10,024,609 44,399,427 54.5
    X 123,194,758 146,604,866 38.5
    X 13,173,093 151,050,290 160.8
    X 13,173,093 154,488,536 167.7
    X 118,624 151,054,252 189.1

    X 850,802 21,936,384 38.9
    X 38,883,698 154,088,731 128.7

  17. So I did the ‘are your parents related’ on gedmatch and the outcome is this:

    Largest segment = 83.1 cM
    Total of segments > 7 cM = 979.0 cM
    Estimated number of generations to MRCA = 1.9


    Largest individual segment of SNPs with greater than 500 matches: 7.0 cM

    Total autosomal SNPs in segments with greater than 500 matches: 183595

    No full matches, only half matches, and looking at the chromosome picture it looks like a good 25-30% is shared. So what does this mean? First cousins? Double first cousins? Randomly shared lines? Trying to figure it all out.

  18. That is the largest I have ever seen from are your parents related. I need to consult with a few other experts but to me it looks like your parents are half siblings or uncle/aunt and niece/nephew. [Correction after consultation likely to be even closer]
    Do you know who your parents are? I will email you since this looks pretty sensitive to me

  19. I still do not quite understand the whole XY thing with half siblings. I am adopted and have a half brother through our birth mother. We came up as a match on ancestry as close family/first cousins. No shock there, we already knew we were related. However, I have a possible half sibling on my birth fathers side who just sent her DNA in. I know my DNA matches up distantly to the family line but this will be the first possible close match I have with my paternal side. If I’m reading right, every father passes down the same X DNA to each of his daughters? So would our test be more conclusive since we are both female?

  20. Catherine,
    You understand it just fine and are reading correctly. If you share a father with another female, your X chromosomes will match for the entire length. Send me the kit numbers once she is tested and uploaded to GEDmatch if you have any doubts.

  21. Hi there,
    I was wondering, could a half-sibling ever be an organ donor to the other sibling? Could their genetics ever be that close?
    The reason I ask, I want to be as accurate as possible for my latest book, where a half-sibling is able to save her brother. If it’s not, I’ll have to go strictly to a fictional ending, but it would be so nice if it could work out.
    I’m glad I found your article today. It was exactly what I was looking for.

  22. Kitty – hello! Very interesting blog and discussion here. It has been most helpful. Based on what I see here and on various other sites, my brother and I seem to share a relatively high number of cMs – per GEDmatch we share 3,586.9 cMs on C1 – C22 (longest block is 281.5 on C1), plus we share 196.1 cMs on the X (which appears to be the full X- right?).

    Can you conclude anything interesting or helpful from this “abnormally” close match?

    Since we both seem to have inherited an identical X chromosome from our mother, can we conclude that we inherited the X chromosome she inherited from her father, rather than the X she inherited from her mother? We do have an X match to a cousin from my mother’s paternal line, so I’m guessing that is the case.

    Any other clues here that can help us in our genealogical research? Thanks tons for your help and insight!

    • Hi Joel –
      Please do a one to one image only compare. If it is solid green and blue like this image – http://slides.com/kittycooper/gedmatch-10#/14 – then you likely uploaded the same kit twice. Or you are identical twins. Or the lab where you tested mixed up your results so that you have the same kit twice.
      That amount of shared DNA means these kits are the same.

  23. Hi Kitty,

    This is my fathers dna and he was given up for adoption. By dna testing we were able to find the mother side and the dna below is show family on the father side. We are trying to see if they are half siblings, cousins etc.
    What can you tell? thank you for your help.

    Largest segment = 54.6 cM
    Total of segments > 7 cM = 234.2 cM
    8 matching segments
    Estimated number of generations to MRCA = 3.0

    • Laurie –
      That is a solid match, likely a 2nd cousin to your Dad (subtract one from that number of generations to get the likely cousin-ship), see http://isogg.org/wiki/Autosomal_DNA_statistics – (it could also be a first cousin twice removed or a few other relationships in that neighborhood).
      Contact them! Tell them the story. Also look if they have a tree posted anywhere. Then look at all their gg-grandparents. Does your Dad match others descended from any of them?
      DNAadoption.com has lots of helpful resources as does DNA detectives on facebook

  24. Kitty – I feel like such an idiot! You are exactly correct. Somehow when uploading my and my brother’s kits to GEDmatch originally, I must have mixed up the file names and uploaded my brother’s kit twice – DOH!

    With our newly uploaded correct kits, GEDmatch shows us sharing a more normal 2,792.7 cMs. It shows “Estimated number of generations to MRCA = 1.2” – shouldn’t this be “1.0”, since we’re full brothers?

    Our X chromosome match is now shown as two large segments of 53.5 cMs and 78.9 cMs, respectively, for a total of 132.4 cMs (so looks like we share approx. 67.5% of the total X chromosome). Are there any conclusions to be drawn from this X sharing level?

    As I mentioned previously, we do know of one certain X match with a second cousin (2x removed) and this second cousin is on my mother’s father’s line. My brother matches this second cousin for two X segments of 24.9 cMs and 20.8 cMs, respectively; and I match this second cousin for the same initial 24.9 cMs segment that my brother matches her, but also match her on a longer 44.7 cMs segment that “contains” (has same start location as) the 20.8 cMs segment my brother matches this cousin on. Any other clues hidden in the way my brother and I match this second cousin?

    Again, thanks so much for your help!


  25. 1.0 would be an exact 50% but most of us are not exact. I share 47% of my DNA with my brother and we are 1.2 also.

    Your X sharing looks normal. Your mother gave you each slightly different X since she recombined what she got from her parents. Sorry no special clues from the X you both share with your 2nd cousin but it may be useful for figuring our other DNA cousin matches.

    One thing about X is that many “sex linked” genes like male baldness, color blindness, and some more harmful ones are carried on the X. So if your mom’s Dad kept his hair you are about 50% to keep yours! See https://en.wikipedia.org/wiki/Sex_linkage

    I see my brother with the same little bald spot that my mother’s Dad had …

  26. Okay, thanks Kitty. Makes perfect sense now. Yes, my maternal grandfather kept all of his hair, and so far so have I, so that’s one plus!

  27. Hi Kitty,

    I have a quick question if you don’t mind. I have just got my dna tested with AncestryDNA. Basically I’m suspecting that my uncle is maybe my father and my dad is actually my uncle. My siblings and I share the same mother and supposedly the same father (but I am suspecting it’s actually my uncle my dad’s brother that is possibly my father) I’m thinking that if I had my sister tested should our X be identical if we share the same dad? if not should it just be segments of the X if my sister father and my father are brothers aka my uncle is actually my dad? Sorry I know this sounds really confusing it is to me too! 😀

  28. Hi Kitty,
    Please can you give me some advice. If two full siblings both share a segment with a first maternal cousin on the yellow segment, would the yellow segment be a maternal match for the sibblings? Also if one of the sibblings have a match to another person on the same yellow segment but this match doesn’t match the other sibling until the segment turns green. Would it be safe to assume this match is to the paternal side as this match doesn’t match the maternal first cousin over this segment?

  29. If two siblings share a segment with a maternal first cousin then it rates to be maternal if it is > 11cM else it might be false (is anyone else from that cousin/s family tested?) although that is unlikely.
    Your second assumption is also likely provided the shared segment is large enough. If it is less than 7cM then it is more likely to be a false match than a good one.

  30. Hi Kitty,
    Thank you very much for your reply. 🙂
    The segments are larger than 11cm.
    Just out of curiosity do you think that endogamy in the population could affect this in this particular situation?

  31. Yes endogamy can affect this. Run the “are your parents related” on all these kits. Although that may not uncover this issue. For example if your maternal first cousin is distantly related to your paternal side, part of the matching segment could possibly come from your paternal side. Unlikely but possible particularly in an endogamous group. If your parents are tested you can use your maternal and paternal phased kit for comparisons. Otherwise go with the assumption you have, but keep comparing to other close relatives.

  32. Dear Kitty,

    My sisters and I have been trying to figure out which share paternal or maternal or both and it has gotten very confusing. My sisters share 194.1 on the X chromosome with each other. I share approximately 117 with each on the X chromosome. My daughter does not shar any X chromosome with either of them. In order to get 14cm with either I had to lower the threshold to about 200 at 2cm. Can you help me understand these relationships? Also, my sisters are supposed to have different fathers. And one more strange part although they share 194 on the X chromosome it says the largest segment is 128.

  33. Hi Kitty again,

    I think I need to give you even more explanations. On the X chromosome one to one match my sisters have a solid blue line all the way across sharing 196.1. The bar above the blue line is yellow. However the X chromosome matrix puts them a little lower at 194.4

    They are supposed to be half sisters to each other and to me. However on the X chromosome I share 117 cm with one and 118 cm with the other and my daughter does not share any part of the X chromosome with either of them. I look forward to learning how we are all related! Thank you!

  34. Gina –
    We use the X to determine which side, paternal or maternal, once we know people are half siblings from looking at the picture of the other chromosomes and the total cM. So look at the image of the other chromosomes when you compare the two of them for the tell tale green bars that indicate full siblings. Also what is the total cM?
    From the X, you appear to be a half sister to them and they appear to have received the same X as each other from their mother, unusual but possible.

  35. Ok- Thanks when you ask for total do you mean autosomal or XDNA

    Sister M- Autosomnal I am 1699.1 total with a 100.4 largest cM t. On X we are 117.8 total cm and 117.8 largest cm

    Sister D- Autosomnal I am 1498.8 total cM with a largest cM72.7. On the X we are 117.9 total cM and 117.9 largest.

    Can you explain why my daughter does not match either on the X chromosome?

    My daughter believes my sisters and I share don’t share the same mother but perhaps our mothers are related.

    Thanks again!

    • Gina –
      It is perfectly possible for your daughter to share no X with her half aunts when you share a mother with them. That would happen if you passed her your father’s X over that 117cM area where you share X with those half sisters.

      The total cM of autosomal is the most reliable predictor for close family BUT half siblings share the same amount of DNA as full aunts or grandparents, 25% or about 1700 cM autsomal. Both your presumed half sisters are a fit.So how much autosomal do they share with each other? Could they be full sisters? Or half sisters sharing a Dad (since they have a full X match)

      Where did you test? The mitochondrial DNA haplogroup will be the same for all of you if you have the same mother. 23andme provides that. The SNPs are tested in the ancestry data but you have to upload it elsewhere to get that information

  36. Could you please help, my son did a DNA test, behind my husbands back, with someone who came on the Internet to say she could be his half sister,
    the result came back 95/5% she also said my husbands brother could be her father, is this percentage high and clear proof.

  37. Hello, would people sharing this DNA definitely be half siblings? Is it possible for these people to be fun siblings? Here is the DNA they share.
    Largest segment = 104.8 cM
    Total of segments > 7 cM = 2,076.6 cM
    50 matching segments

    Thank You

  38. I was adopted I did the 23 and me and also just uploaded to Ged website ..There is one person that comes up on both sites ..it says we share 9 segments …500snps….total cm 2232and cm7.0 and 57.5 largest and MT) T2..And of course I’m lost with all this .Can someone help me …Thank you

    • Annette
      What is the total cm – 223.2 or 2232? I recommend you read everything you can. It takes a while to understand the significance of these matches and the various numbers. Go to DNAadoption.com and read their materials.
      Look at the charts here
      I will email you and have a quick look

  39. Quite a few of the posts right above are what I am currently dealing with. I love your comment that the DNA doesn’t lie. I uploaded my info to GEDMatch from Ancestry.com. Two other people that I had never met, but had shown a connection with on Ancestry, did the same thing. I would love to give you more details via email and ask your opinion if you get the chance (even though I am fairly certain I now know the truth and I am OK with it!:). I will be looking for the book Stranger in My Genes, for sure! Love this blog. So happy I found it. Thank you!!!

    • TK –
      if you have uploaded to GEDmatch, I am happy to take a quick look. Otherwise I just do not have the time to do this for free. I will email you.

  40. Ms. Kitty,
    Me and a gentlemen who I’ve always thought was my full brother took the ancestry DNA test. I uploaded the results to GEDMatch and we only share 1663.3 cMs. There were a few high numbers like 119, and a few 90’s in some areas but in your opinion is he my half brother and is it even possible that he could be my full brother?

    • Sounds like you are half siblings, sorry, but he is still your brother

      Look at the graphic comparison at GEDmatch. Half siblings will have no big green bars as in the 2nd image above. The first image is full siblings with the green bars

      • Thanks Ms. Kitty,

        I checked again and there are about 3 or 4 small green spots that are as big the 2nd largest green area in your Chr 2 posted for your full sibling. I was trying not to believe it but I guess its true. Thanks so much for your help!!

  41. That 1136 cM is most likely a first cousin or a half nephew. What is your age difference?

    Yes his mother’s father’s side is where it will be from the lack of other matches. You could be his mother’s half sibling on her father’s side for example. I take it she has not tested or is not available?

    So try and get relatives on his mother’s father’s side to test. Does his mother have any siblings? Do they have descendants, etc

  42. If I were to post results of a dna test for half-siblingship, could you tell me what the probability is? It says 98.7% and I don’t see how that is possible since we only are looking for half siblingship. Please help. Im “dna comparison” illiterate lol.

    • Probably not, my expertise is with autosomal tests from companies like ancestry.com or 23andme or FTDNA not with tests for specific relationships

  43. Hi Kitty,

    We just got the results from my Mom’s and her Sister’s DNA test at FTDNA. We are trying to confirm whether they are full sisters, half, or cousins. The results came back with a match of 2,445 cM, X match at 195 cM, longest block 137. They said possibly Full siblings, Half siblings, Grandparent/grandchild. Would it be possible for you to look at the chromosome browser and tell if they are indeed full sisters? Thank you!

  44. I’m trying to figure out if my half sister’s father (we have the same mother) is my fathers half brother or not. The question of who her father is has been debated in whispers for a long time This along with the fact that she has about 3-4% more Native American than I do and, as far as my knowledge goes, our mother and my father as well as his half brother are the only ones with native in them, makes me wonder what our DNA is telling us.
    We both did the Ancestry DNA test and it matched us as cousins. I uploaded both to GEDMATCH and we share 59 segments and 2064.2 cM with our highest numbers being 93.3, 90.5 and 70.1.
    Would this show us as “sister cousins” or as half siblings?

    • Katelyn,

      That much shared DNA confirms half sibling or 5/8 if her dad is your dad’s half brother unless she could be your aunt or niece … See http://isogg.org/wiki/Autosomal_DNA_statistics

      What does the image only view of your comparison with your half sister show? Any of those green bars? You will have some if her father is your father’s half brother.

      You can confirm the half brother by testing 1st/2nd cousins related to him and seeing if they match your sister

      • The image only view shows a lot of yellow, green and blue. It doesn’t show any big bars of red but there are bits. It does however have several large bars of green, some taking up about 1/4 to 1/3 of the bar.
        Our 21 and 22 chromosomes are only yellow and green for full and half matches with the bar of green being more than half and the entire bottom is blue

  45. The green bars, if they are large, suggest that either you are full siblings or that your fathers are related. I will email you.

  46. I believe I was scammed. I submitted my follicles for a DNA test. It took almost a month. I think the viability issues arise there. I was told I have a daughter. I believe the mother may have submitted the daughter half-brothers DNA. Could this come up as a match?

    • My expertise is with autosomal DNA tests not anything involving a follicle? If you and the daughter tested at Ancestry.com, familytreeDNA, MyHeritage, or 23andme the results would be clear. DNA does not lie. Otherwise I cannot help. For all those tests, the sex of the testee is clearly indicated by whether there is an XY or an XX.

  47. Jim without knowing where the test was done I cannot give you an answer. If this was an autosomal DNA test then the amount of DNA for parent child is unmistakable and a half sibling would share way less DNA. Autosomal tests like Ancestry use about 700,000 markers so are far more accurate than drugstore paternity tests

    • Ok, I think you answered my question. There is no way a half-sibling could be mistaken for a parent. I appreciate your answer though I was hoping for another.

      It has been a long ordeal since 2012. Finding out on Father’s Day about a 12 yo child. Then finding out she was given up for adoption. Then finding her. Then drama and no contact for 2 years. Then more drama…

      I keep trying because its the right thing to do because of the DNA test. I was hoping to be released on the possibility there was DNA fraud. But I guess not.

      Ill have to wait a few more years to see if she is ready.

      Thank you again.

  48. Hi kitty,
    I’ve been reading everything on your blog for over a month, but not sure I understand…. my uncle was rumored to have a different father, he shows relation to me as 1st cousin 902 cms & 43 dna segments
    So I had his sister tested and relatons to me is 1st-2nd cousins 852 cms across 40 dna segments so my Aunt has a different father as well?
    Do I understand this right?

    • It sounds like each child had a different Dad. The 1600 some cMs is half siblings and your numbers are half niece.
      Perhaps your grandmother’s husband was infertile so she used donors …

  49. What would be the expected cM shared between half siblings if: Brother B and sister S have child A and then Sister S and Father F have a child C. What would the expected shared cM be for Child A and Child C?

  50. 1080/43 Is it at all possible that this could be a half sibling? It is definitely not an aunt/niece match . Ancestry has said first cousin but when I called them they said they can not distinguish between a first cousin or a half sibling. Is there any way this could be a half sib or is it def too low.

  51. The girl that matched 1080/43 would either be a first cousin as her father would be my fathers brother. The other possibility as that we share the same father but different mother. There are basically two brothers involved and I am adopted.

  52. f.,

    Using Ancestry DNA testing, is it possible for a half-sibling and/or half-niece to appear as a 3rd-4th cousin instead of a sibling—given the explanations you have given under this topic concerning full- vs. half-siblings? Thanks!

  53. Hello!

    I supposedly share the same father as a newly discovered sister. We both did a 23 and me test and it said there was no overlapping DNA matches. Is it possible 23 and me got it wrong? I have adoption paperwork and a parental rights termination record with her father’s name listed as my biological father. We even look similar.


    • Luke –
      Something is wrong. Most likely he is not your Dad. Looking similar is not the final word. DNA does not lie.
      You could contact 23andme and ask if there is any possibility that there was a mix up in the lab, but that is unlikely.
      Best though would be to both retest at Ancestry or another company. Or perhaps just you retest, since you will probably want to search for your biological Dad. You can then compare your new test with hers at a third party site like GEDmatch.com

  54. One more question, I promise. Do you see a pattern with Ancestry DNA tests being off with their numbers. Is it possible they could be off by 200 cm? If they are lower than say 23 and me then I think it would be possible to be in the half sib range?

    • Fran, ancestry removes a small amount of matching Data that it considers “population” specific so that is possible but very unlikely to be so large. Please upload to GEDmatch and send me the two kit numbers via my contact form and I will look for you

  55. Kitty – I have another half vs. full sibling problem for you, if I may.

    According to GEDmatch, two sisters – one who tested at Ancestry and the other who tested at FTDNA – have an autosomal match of approximately 2900 cMs (longest block is 188 cMs), and have a full X-match at 196 cMs. However, the sisters have been told all their lives that they have different fathers – but that the two fathers were brothers. Is it possible, based on their DNA results that they could actually be “half” sisters, whose fathers were brothers, rather than full sisters who share the same father?

    When I saw the sisters’ results with the relatively high autosomal match of 2900 cMs and their full match on the entire X chromosome, I thought they absolutely had to be full sisters with the same father, notwithstanding what the sisters have been told. However, I’ve read recently that brothers can sometimes be identical X matches to each other (though I haven’t seen it yet).

    So, could these two sisters have received the exact same X chromosome but from different fathers, who were brothers, making the two sisters have a full X match (196 cMs)? Or, does the sisters’ “high” autosomal match of 2900 cMs, plus their full X match, mean the different fathers who were brothers scenario is really not possible?

    Oh, the fathers were NOT identical twins, but they were full brothers.

    Thanks for your help!

    • This certainly sounds like they are full sisters. I will email you for their kit numbers at GEDmatch but between the X and the high number of shared centimorgans, they are full sisters.

      Any chance the two brothers could test?

  56. Hello Kitty,

    My mom has a close family match on ancestry.com with whom she shares 1,659 CM and a first cousin match 872 Cm. My mom doesn’t know her biological father. What kind of match would show as a close family match?

  57. Hi kitty,

    I recently found a half sister on ancestry And uploaded the results on GEDmatch. The number seems very high though. Our shared centimorgans are 2,214 over 34 segments. Our largest shared segment is 182.8 then 146, 141, 131, 124, 114. There are many in the 70 to 90 range also. Does this seem like a normal possible range for half sisters?

    • Erin –
      That is high for a half sibling but almost within the range. I have seen other half siblings this high. If you come from a group that is at all endogamous, then these larger numbers are more likely.
      But look at the two images above. Do you have any of those green bars? If you have quite a few of them, then you are full siblings, otherwise with none or maybe one small one, you are halves. Of course the other possibility is that the parents (fathers?) you do not share are related!
      If you are still not sure, use my contact form to send me the two kit numbers but it may be a few days before I can look.

  58. I just got my results back and I’m having troubles understanding them is a 45% mean that me and the other one tested in the half sibling test half siblings ot none at all

  59. Does a 2344 cMs reading indicate full or half siblings? Seems to be on border line. Same mother just question about father.

    Thank You.

    • Jeff it could be either. Where are you tested? Ancestry removes population specific segments so it might be full siblings if tested there.

      More likely these are half siblings but compare them at GEDmatch and look for the green bars that indicate fully identical segments (FIRs). Full siblings will have quite a few of those.

      Also 23andme now shows those FIRs too.

  60. Both of us were tested by Ancestry.com DNA services. I live in Calif and my sibling in GA, However, we were both born in Pennsylvania but grew up in So. CA.

    • OK Jeff, you could be full siblings or half. You need to compare at GEDmatch in order to see which.
      Unless you each have close paternal and maternal relatives tested, then you can tell by whether they all match your sibling.

  61. “I have not been able to find this tool in the new 23andme experience, so I guess it is gone. If anyone finds it let me know.”

    I suppose you are aware that this feature is back now. But quirky.

    Back for a month or so. At first it only worked on your siblings. Now you can apply it to any pair of siblings or double cousins etc.

    It tells you how many Full (called Completely) Identical bars you have, but then the chromosome browser misses about half of them.

    Underneath is a very nice chart that is very accurate. And better than GedMatch since it includes the Start and Stop points for the Full/Completely Identical segments.

    • Yes I put an update in this post and I guess I need to blog about it sometime soon and wonderful that it shows the numbers!

      • Hi Kitty,
        Through ancestry I think I have found a possible half sister. We have both uploaded our DNA to gedmatch. We believe we share the same father. I’ve looked at your 1/2 sibling test and it seems similar but would like to have someone with experience to check. Can you help? Ps my father is deseased.

    • A 3/4 sib will have some FiRs but nowhere near as many as full sibs. The shared cMs will be around 2100. If you have such a case, send me the kit numbers via my contact form and I will give an opinion.

    • Sharon
      The children of double first cousins will look more like first cousins than the double 2nd cousins that they actually are

  62. MsKitty,

    I just entered the DNA kit numbers for my wife’s and her sister into GEDmatch. Up to a few moments ago we were positive that they were half siblings. The GEDmatch analysis listed on almost every chromosome matching FIRs, lots of green. We used Ancestry.com for the DNA raw data. Could there be the Ancestry.com data be giving a false reading or are they full siblings?

    • Sounds like they are full siblings after all! How many cMs?

      Send me the kit numbers and back story via my contact form and I will have a look later

  63. I just forwarded info (kit numbers etc) you requested on your contact form.

    Thank You


    FYI 2585 cMs over 60 dna segments

  64. We feel a little confused….. My husband did Ancestry DNA (he’s adopted, was looking for heritage, didn’t even know they gave matches!). He matched as close relative 1721cm’s over 51 segments with a woman. We can’t seem to pinpoint what the relationship is. (Not his birthmom) The second closest match is this woman’s half sibling (they have the same father), which the numbers to my husband are only 670cm over 20 segments. The high match seems to think my husband is her half sibling…wouldn’t the 3 of them have around the same #’s if they all had the same father?

  65. Lisa, that is very unusual. The numbers suggest that she is his aunt and her half brother is his half uncle.
    If they are all on gedmatch, send me the kit numbers via my contact form and I will look delicious more

  66. I found your blog while trying to find information on how to interpret my Ancestry DNA results. I have an estimated “first cousin” match that shares 1300 centimorgans across 47 segments. Our trees don’t match at all. Everyone tells me that that’s a rather high match for first cousins. I was wondering if it points more toward half-siblings? The match is only one year younger than I am, and I cannot see how he could be an uncle or nephew. There’s no way he could be my grandparent or vice versa. I feel completely blown away by the results I got. We also share a “2nd cousin” match (243cm across 11 segments). Half of my family tree no longer makes sense, and my heritage wasn’t at all what I thought it would be. Could someone help me make sense of this? I haven’t been able to get any responses when trying to message them.

  67. Amber –
    I am so sorry for your shock. This happens more than you would expect. Your parents are still your parents even if one of them is not biologically so.
    Or perhaps your half sibling (if he is) is the one to be surprised.

    Please join DNA detectives on Facebook and ask for help and/or the mailing list for DNAadoption.com on Yahoo.
    I will email you privately

  68. Kitty

    I have a question regarding 1/2 sib. and 1/2 aunt/uncle match.

    Seems most of these post here share a lot of DNA for their 1/2 sib matches… makes me wonder if I am on the right track or not. Both my mom and dad were adopted children and I recently got a new match on ancestry DNA and the match to my mom says close family-1st cousin (Predicted relationship: Close Family
    Possible range: Close family – 1st cousins
    Amount of Shared DNA
    1,351 centimorgans shared across 38 DNA segments)
    and same person to me says
    (Predicted relationship: 1st Cousins
    Possible range: 1st – 2nd cousins
    Amount of Shared DNA
    809 centimorgans shared across 30 DNA segments)

    Is it possible that this person would be my mom’s 1/2 sibling and my 1/2 aunt with these numbers? This persons DNA is not on gedmatch.

    On a separate note: Gedmatch shows a tie between on of your tests, my mom and me: (one to one with my mom)
    Largest segment = 34.4 cM
    Total of segments > 7 cM = 134.6 cM
    7 matching segments
    Estimated number of generations to MRCA = 3.4

Leave a Reply

Your email address will not be published. Required fields are marked *