Shared DNA (cM) Calculator

What a centimorgan measures

DNA testing companies report matches in centimorgans, cM, a unit of genetic length based on how often a stretch of chromosome is broken up by recombination when it is passed from parent to child. The 22 autosomes add up to roughly 3,400 cM per copy, and you carry two copies, one from each parent, about 6,800 cM in all. A parent and child share one full copy, about 3,400 cM. Each generation further back halves the expected amount on average, which is why shared cM is the main evidence for estimating how two testers are related.

Expected values by relationship

Theory gives the average. Grandparents, half-siblings and aunts or uncles all expect 25%, about 1,700 cM. Full siblings are a special case: they share half their DNA on average, but because some segments are identical on both copies, testing companies report about 2,550 cM, three-quarters of 3,400. First cousins expect 850 cM; first cousins once removed and half first cousins, 425 cM; second cousins, first cousins twice removed and half first cousins once removed, 212.5 cM; third cousins, 53 cM. Large surveys of testers, such as the Shared cM Project, find slightly higher averages for distant relationships, around 229 cM for second cousins and 73 cM for third cousins, because the smallest shared segments are not reported and so the matches that are reported skew upward.

Why one number fits many relationships

Each parent passes on a random half of their DNA in large chunks, so the real amount varies widely. Full siblings range from about 1,600 to 3,500 cM, and second cousins from almost nothing to nearly 600 cM. A figure of 220 cM is closest to the theoretical value for a second cousin, but it is equally consistent with a first cousin twice removed, a half first cousin once removed, a half great-great-aunt, and with the low end of a first cousin once removed. The calculator ranks the relationships by how close their expected values are, but it cannot decide between them. That needs evidence: ages, shared matches who are known relatives, the longest shared segment and the family tree.

Common pitfalls

A common mistake is to treat a match’s predicted relationship from a testing company as a fact. It is a probability, and the right answer is often the second or third suggestion. Endogamous populations, such as communities that have intermarried for centuries, share extra DNA through many lines, so cM figures overstate how close the relationship is. Very small matches, below about 7 cM, are often false matches rather than real shared ancestry.

How to use it

1. Choose whether you are starting from a cM figure or from a relationship.
2. Enter the total shared cM from your match list, or choose the relationship.
3. Read the closest expected relationships and how far your figure is from each.
4. Narrow the candidates with ages, shared matches and documentary research.