A question has been raised privately that needs to be responded to here. In essence, the question is "If two participants share the same distant ancestor according to their research, but the DNA results indicate that they are not related, could the DNA sample somehow have been corrupted? And, what does it take to get the lab to do a retake?" Quoting from: http://www.familytreedna.com/faq.html "The chance that a match does not exist due to infidelity or unreported adoptions occurs 2%-5% of the time per generation. For families trying to do family reconstruction, it is prudent to test 2 different known male cousins from each different branch. "In cases of unexpected results, FTDNA will re-test at no charge to confirm that a lab error is not an issue. In case that a retest is done and the results are confirmed, a $50 charge will be applied." So to answer the second question first, a participant who believes that the lab made an error in testing his sample can request that it be re-tested. Family Tree DNA will do so. If an error is found, there is no charge for the re-test. If they get the same results the second time, there is a $50 charge for the extra testing. Before someone rushes to have a sample re-tested however, I recommend that they reexamine their assumptions about their assumed family tree. I would asking the question at each generation, "Do I know with absolute certainty that the connection I have assumed did in fact exist?" Let me explain what I mean with an example. I was researching a friend's line (not Bowles) and we believed that the census evidence clearly linked his father with a family that came from Canada to New York. The family names and ages and the locale of where his father grew up all matched this family. Since my friend's father had run away from home as a teenager and had never told his son (my friend) anything about his grandparents, we had only the census evidence to guide us. Believing that we had made the correct link, we continued to investigate this Canadian family and traced it accurately back several generations. Just when we were patting ourselves on the back for finding his great-great grandparents, we discovered that our original assumption was mistaken. We found another census record with a weird spelling of his surname that clearly links him, not to the Canadian family, but to one in Boston. As I see it, there are five reasons why two participants who claim the same distant ancestor would receive radically different DNA results: 1. There was an adoption somewhere in the family tree. 2. There was infidelity somewhere in the tree. 3. A mistake has been made in connecting families within an assumed tree. 4. The lab made a mistake. 5. The sample itself was corrupted before the lab received it. It is pretty difficult, but not impossible, to track down instances of #1 or #2, especially if they occurred several generations back. A thorough re-examination of one's assumptions about his tree may be the first step to take. Before you request that the lab re-test your sample, consider carefully if there is any way that the sample could have been corrupted before they ever received it. Was the right person tested? Did anyone else touch the end of the swab stick where the DNA was collected? The chances of this happening are very small, but must be considered. If you've re-examined your assumptions about your family tree carefully and still conclude that you belong on a tree with someone who got different test results, then consider having the sample re-tested. One last thought... is it likely that the lab would misread a sample so badly that two samples that should have been the same would end up with mismatches on 10 or 12 markers? I could more easily understand them misreading one or two markers, but not 10. I guess I've muddied the waters even more, but all of this should be considered. Charlie Hartley Bowles DNA Project Administrator [mailto:hartley@iglou.com]