Centimorgans Explained: The Unit of DNA Sharing

If you have ever looked at a DNA match list, you have seen a number labeled “cM” next to each relative. It is the single most important figure in genetic genealogy, and it is also widely misread. A centimorgan is not a count of matching letters or a percentage - it is a unit of genetic distance, and understanding it changes how you read every match.

What a centimorgan measures

A centimorgan (cM) measures genetic distance by the likelihood of recombination. Recombination is the shuffling that happens when your DNA is passed down: chromosomes swap segments before being handed to the next generation. Two positions that are far apart on a chromosome are more likely to be split up by that shuffling; two positions very close together tend to travel together.

The centimorgan captures that likelihood. It is a measure of how prone a stretch of DNA is to being broken up across generations, rather than a physical length in base pairs. In consumer DNA testing, it is used to size the shared DNA segments between two people - the runs of DNA they inherited intact from a common ancestor.

More shared cM means a closer relationship

The core intuition is simple. Close relatives inherited large, unbroken stretches of the same DNA, because there have been only a few rounds of shuffling between them and their shared ancestor. Distant relatives share smaller, more chopped-up pieces, because many generations of recombination have whittled the shared segments down.

So the more total cM two people share, the closer their relationship tends to be. A large shared total points to a near relative; a small total points to a distant one. This total, added up across all the shared segments, is the number your matching service reports.

The numbers that anchor the scale

Two figures give the whole system its scale:

• The total autosomal genome is around 6,800 cM.
• A parent and child share about 3,400 cM - roughly half of a person’s total.

That parent-child figure is the reference point everyone else is measured against. It makes sense: you inherit exactly half of your autosomal DNA from each parent, so a parent-child pair shares almost precisely half the genome, which comes out near 3,400 of the roughly 6,800 total cM. Full siblings share about half on average too, but through a different, more variable mixing, so their totals scatter more.

Why one number is not one answer

Here is the part that trips people up. Relationship predictions use shared-cM ranges, and those ranges overlap. A given amount of shared DNA does not point to a single relationship; it points to a set of possibilities.

For example, a mid-range cM total could be consistent with a half-sibling, an aunt or uncle, a grandparent, or a niece or nephew, because all of those relationships can produce shared totals in the same neighborhood. The DNA alone cannot separate them. That is why a responsible tool reports a cM total alongside a list of candidate relationships rather than one confident label.

The takeaway: treat a shared-cM figure as narrowing the field, not naming the answer. You combine it with ages, known family structure, and sometimes additional relatives to resolve which possibility is real. Our chart of DNA shared with relatives lays out how those ranges stack up.

Reading matches with this in mind

Once you see cM as genetic distance rather than a simple score, match lists become far more honest. A high number is a strong signal of closeness; a low number is genuinely ambiguous and easy to over-interpret. The mechanics of how services turn your genotypes into these shared segments are covered in how DNA relative matching works, and you can compare two of your own files directly with private comparison.

This article is educational and is not medical advice.