In 2002 I volunteered at the national genealogy workshop where a university prof was collecting DNA & pedigree charts for a theory to see if DNA could be used to research genealogy. Well a local company hired him away from the university and in time his study went on line. A fellow I had been emailing with for 10 years called me and said, "You know how you've wondered if your family changed the surname from O'Brien to Bryant. Well cousin..." Come to find out he and I matched 100% DNA at 43 markers. He lives in Buffalo, NY, and his grandfather came to U.S. in 1903, mine came to Canada about 1830, then to U.S. in late 1850s. Our common ancestor lived about 1700 in the 8th-10th generation. Sadly the records in Ireland for this time period are very scarce. But for me it proved my surname change and origin of the family in Ireland. He has gone on in the clan surname project also at Family Tree, to find four others who are with in 1 or 2 markers difference from us, including another at 100% at 37 markers. These others O'Brien ancestors came from 1-3 miles from the same western Co. Clare town of Killarnen from one another. So for all of us it was money well spent, yet most I've read didn't have such success. On O'Brien Clan director in CA did the test & found out he had the surname but not the DNA, he resigned. Other DNA surname projects have help to determine traditions whether they be true or false. Good luck in the decision. I'd chose 67 or 87 marker test. The more markers tested the more refined the test/results. In matches with people there is probably a relationship between 1-4 marker variation, but 5 markers off indicates no relationship. The less markers off the closer the relationship.