Revised August 2026.
DNA matches inform but may pose questions that are hard to answer. This page has references to issues that may arise or may need consideration when using DNA results.
It is very useful to have tests on as many of your close family as you can. By close family this starts with as many of your parents and grandparents as you can; followed by as many brothers and sisters. Then I would enlist as many 1st cousins as possible.
Each individual inherits DNA from their parents practically equally, but which parts of the DNA passed on from each parent is random. Looking at the sets I have there are plenty of examples where one sibling has strong matches to DNA from a great-grandparent line and the other has none. In the case of my nieces this is even more marked and they have DNA matches, to persons descended from ancestors further back, that I don't have. This is often useful to confirm possible ancestors where the documentary evidence is not conclusive. The scope of this variation is best seen by the results available by painting DNA segment information on DNA Painter as outlined on the 'Test Information' page.
Where matches show strong relationships but the documentary evidence suggests otherwise are often referred to as Non-Paternal or Non Personal Events (NPE). So far such DNA pointers allow people to find their biological relatives, or at least suggest possible ancestors but such matches may cause some persons upset. See the page Unexpected Matches below where these issues are set out and two links to articles that discuss these, including one on adoptees.
The most useful tests are those where the person tested has also provided a family tree to which the test is attached. This should always be checked against actual records as errors are easily made. Many people just copy from one tree they find on-line without so doing so perpetuating errors and assumptions. The commoner the surname the easier it is to make these kind of errors. As an example of the more absurd, I saw a person with a baptism in 1735 and a marriage in 1845! On the upside, DNA matches have led to the discovery of tree branches not known or not explored. A fine example is a match to Peter Waterman at 43cM on 3 segments, who led me to discover his links to Dorset when most of his family came from Northern Ireland, yet he is descended from no less than four families from Dorset.
Finally you are likely to find you have sets of matches that form a group but have no documentary evidence for its origins. I have several of these and they are discussed in the DNA results for specific areas. An example I have are to the the Todd/Bettles families from Northamptonshire but whose trees are all from the Carolinas in the US going back to the 1700s. Another example are from Newfoundland families that actually came from Dorset and Hampshire back to the late 1600's in some cases. I have no progress with the Todd/Bettles families while the Newfoundland families show a confusing mixture of cross inheritances with a number of families originating from Poole and surrounds.
This site has the ability to show how tested persons relate to one another and so visualize what a match on a genealogy web site like '2nd cousin 2x removed' actually means! This is seen on any of the descendant charts for a family. At the top and bottom of each chart is a legend showing the meaning of the colours for each person's status in the family; the one marking the tested persons shows yellow. Hovering over the DNA symbol in the real chart shows how many tested persons are on that chart.