Free calculator

DNA Percentage Calculator

Convert between shared percentage, centimorgans and relationship in any direction, and see why the observed value almost never lands on the textbook number.

Runs in your browser No account needed No data stored Dataset v1.0.0-simulated

Convert in any direction

Total used: 6,970 cM
Dataset v1.0.0-simulated · provisionalLast reviewed 2026-09-14Data & sources

Percentage by relationship

RelationshipExpected %≈ cMUsual range
Identical twin 100.00% 6,900 6,700–7,100
Parent / child 50.00% 3,485 3,485–3,485
Full sibling 50.00% 2,594 1,916–3,154
Half sibling 25.00% 1,718 908–2,546
Grandparent / grandchild 25.00% 1,733 652–2,844
Aunt/uncle — niece/nephew 25.00% 1,705 966–2,447
Half aunt/uncle — half niece/nephew 12.50% 841 225–1,668
Great-grandparent / great-grandchild 12.50% 851 166–1,773
Great-aunt/uncle — great-niece/nephew 12.50% 835 259–1,728
First cousin 12.50% 829 266–1,516
Half great-aunt/uncle — half great-niece/nephew 6.25% 409 25–1,074
2nd great-grandparent / 2nd great-grandchild 6.25% 416 7–1,196
2nd great-aunt/uncle — 2nd great-niece/nephew 6.25% 406 34–1,055
Half first cousin 6.25% 409 23–1,057
First cousin once removed 6.25% 402 39–957
3rd great-grandparent / 3rd great-grandchild 3.13% 198 7–753
Half first cousin once removed 3.13% 196 7–734
First cousin twice removed 3.13% 193 7–673
Second cousin 3.13% 194 7–677
First cousin three times removed 1.56% 92 7–442
Half second cousin 1.56% 93 7–462
Second cousin once removed 1.56% 92 7–443
Half second cousin once removed 0.781% 47 7–322
Second cousin twice removed 0.781% 46 7–284
Third cousin 0.781% 46 7–286
Second cousin three times removed 0.391% 27 7–223
Half third cousin 0.391% 28 7–240
Third cousin once removed 0.391% 27 7–213
Third cousin twice removed 0.195% 20 7–161
Fourth cousin 0.195% 19 7–180
A descending bar chart showing shared DNA halving at each step: parent 50 percent, grandparent 25, great-grandparent 12.5, and so on down to under 1 percent.
Each generation halves the expected share. By the fifth generation the expected contribution is small enough that chance starts to dominate.

Why percentages mislead

A percentage feels precise, which is exactly the problem. The expected value is an average over many pairs of relatives; your pair is one draw from a distribution. Reporting "we share 3.1%, so we are second cousins" treats a single sample as if it were the mean. The honest sentence is "3.1% is consistent with second cousins, and with four other relationships".

Common questions

Why do different sites give slightly different percentages for the same relationship?

Because they divide by a different total. Testing companies measure between roughly 6,700 and 7,100 cM of autosomal DNA depending on the chip and the reference map they use, so the same shared segment converts to a slightly different percentage on each platform.

Is 50% from each parent exact?

For autosomal DNA, yes — that is the one number with no variation. Everything further back is a shuffle, which is why grandparent percentages are rarely a clean 25%.

Why is my sibling not exactly 50%?

Siblings inherit different halves of each parent's DNA. On average that works out to about half shared, but the shuffle means it can land meaningfully above or below. Companies also differ in how they count regions where both chromosomes match, which changes the reported figure.