Beginner

Shared cM Explained: What Your DNA Match Numbers Really Mean

Every DNA match comes with a number of centimorgans. What a cM is, what the amount of shared DNA says about a relationship, and why the predicted relationship can be wrong.

3 min read

Every DNA match list shows a number like "212 cM across 9 segments." That number is the most important piece of information about a match, but it's also one of the most confusing. Here's what it means and how to use it.

What is a centimorgan?

A centimorgan (cM) is a unit for measuring stretches of DNA. Technically, it's based on how likely a stretch of chromosome is to be broken up when DNA is passed from parent to child, but you can think of it simply as a measure of length. Each person has roughly 7,000 cM of autosomal DNA to compare, though the exact total depends on the testing company.

When two people share a matching stretch of DNA, inherited from a common ancestor, that stretch is a segment. The total shared cM adds up all the matching segments.

More cM means a closer relationship (usually)

Because you inherit half of each parent's DNA, the amount you share with a relative roughly halves with each generation of separation. Approximate averages, from the Shared cM Project (a large crowdsourced study by genetic genealogist Blaine Bettinger):

  • Parent or child: about 3,485 cM
  • Full sibling: about 2,613 cM
  • Grandparent, aunt or uncle, half sibling: about 1,700 to 1,760 cM
  • First cousin: about 866 cM
  • First cousin once removed: about 433 cM
  • Second cousin: about 229 cM
  • Third cousin: about 73 cM
  • Fourth cousin: about 35 cM

Try our relationship calculator for any relationship.

Why one number fits several relationships

These are averages. The real amount varies a lot, because inheritance is random. Two second cousins might share 100 cM or 350 cM. And different relationships overlap: a match sharing 900 cM could be a first cousin, a half aunt, a great-grandparent or a great-aunt. Your match list's "predicted relationship" is an educated guess, not an answer.

The best way to explore the possibilities is a tool like DNA Painter's Shared cM tool, which shows every relationship consistent with a given amount and how likely each is. Then use ages, trees and shared matches to decide which fits.

Age and generation clues

Ages narrow it down. A match sharing 1,750 cM who's 70 years older than you is likely a grandparent; one who's your age is likely a half sibling. A 433 cM match a generation older than you is more likely your parent's first cousin than your own half first cousin. Always ask: which relationships fit both the DNA and the ages?

Segments matter too

The number of segments and the size of the largest one help too. Close relatives share a few large segments. Distant relatives share fewer, smaller ones. A match sharing 40 cM in one big segment is more reliably related than one sharing 40 cM in five tiny pieces.

Small matches: handle with care

Matches under about 20 cM are tricky. Many are real but very distant, with a common ancestor too far back to identify. Some are "false" matches, with segments that match by chance rather than by inheritance from a recent ancestor. Some populations, such as Ashkenazi Jewish, Acadian and other communities with many intermarriages over generations, share lots of small segments from many ancestors at once (called endogamy), which inflates shared cM and makes relationships look closer than they are.

Focus your research on matches above 20 to 30 cM unless you're working on a specific problem with good tree evidence.

When the numbers don't fit

  • Higher than expected: you may be related in more than one way (common in small communities), or the relationship is different from what you assumed.
  • Lower than expected for a close relative: a half relationship (half sibling instead of full, half cousin instead of full) is a common explanation. See Unexpected DNA Results.
  • Different companies give different numbers for the same pair of people, because they test and calculate differently. That's normal.
Try this: take your top five matches, look up each one's shared cM in a relationship chart, and list the possible relationships. Then use their ages and trees to choose the most likely. Next, read Shared Matches and Sorting Your List into Family Lines.