How Much DNA You Share With Each Relative

When people first look at a DNA match list, the same question comes up: how much DNA should I share with this person, and what does that tell me about our relationship? There is a rough pattern, and once you know it, match lists start to make sense.

The halving pattern

Every generation of shared ancestry roughly halves the DNA you have in common. That single idea explains most of the chart:

• Parent and child share about 50 percent. This one is fixed - you inherit exactly half of each parent’s DNA.
• Full siblings also share about 50 percent on average, but here is the twist: you and a sibling inherit different halves from each parent, so you share different segments, not the same 50 percent.
• Grandparent and grandchild, aunt or uncle with a niece or nephew, and half-siblings all cluster around 25 percent.
• First cousins share about 12.5 percent.
• Second cousins share roughly 3 percent.
• Beyond that, the number keeps roughly halving with each step outward, shrinking toward the point where shared DNA becomes hard to detect at all.

Notice that several different relationships can land on the same percentage. A half-sibling, a grandparent, and an aunt can all sit near 25 percent, which is why a raw percentage alone rarely pins down an exact relationship.

Percentages and centimorgans

DNA sharing is often measured two ways. A percentage tells you what fraction of the genome two people have in common. Centimorgans (cM) measure the genetic length of the shared pieces. Match tools usually lean on centimorgans because they capture how much shared material there is in absolute terms.

For scale, a parent and child share about 3,400 cM - close to the maximum two different people can share. As relationships get more distant, the centimorgan total drops in step with the percentage. If you want the full story on what that unit actually measures, see our guide to what centimorgans are.

Why these are averages, not guarantees

The single most important thing to understand is that every number here except parent-child is an average across many people, not a promise for your specific case.

The reason is recombination - the shuffling of DNA that happens when it is passed down. Because the shuffle is partly random, two people in the exact same relationship can share noticeably different amounts. Two sets of first cousins might share somewhat more or less than the textbook 12.5 percent, and both are completely normal.

This variation grows as relationships get more distant. Close relatives fall in a fairly tight range, but out at the second-cousin level and beyond, the possible spread is wide. It is even possible for a real, distant cousin to share so little detectable DNA that they do not appear as a match at all.

Reading a match list wisely

A few practical habits help you avoid over-reading the numbers:

• Treat a shared amount as a range of possibilities, not a single verdict. A given centimorgan total is usually consistent with several different relationships.
• Use shared matches and family knowledge to narrow things down. The DNA number is one clue among several.
• Expect surprises at a distance. The further out you go, the more the actual sharing can drift from the average.

For a fuller picture of how services turn shared segments into a match and an estimated relationship, see how DNA relative matching works.

Keeping it in perspective

The relatives chart is one of the most useful mental models in genetic genealogy, but it works best as a guide rather than a rulebook. The halving pattern gives you strong expectations; recombination adds the scatter around them. Held together, they let you interpret a match list with confidence without pretending the numbers are more precise than they are.

If you want to explore shared segments between two of your own files without sending anything to a server, you can do it privately with on-device DNA analysis, where your raw data never leaves your device.

This article is educational and is not medical advice.