Full siblings inherit different random halves of each parent's DNA, because each parent's chromosomes are cut and swapped (recombination) before being passed on. On average siblings share about half their DNA, but ancestry estimates are built from which segments you received, so the numbers differ, especially for smaller regions.
That is normal. A large mismatch is rarely a lab error, and a few unusual cases are worth understanding.
How the shuffle works
Each parent has two copies of every chromosome, one inherited from each of their own parents. To make an egg or sperm cell, the two copies swap segments, and the cell receives one mixed copy. Each child therefore gets a different patchwork of segments from each parent's grandparents. Siblings are drawn from the same four grandparents, but not the same pieces.
On average, full siblings share about half of their DNA, according to the International Society of Genetic Genealogy (ISOGG). The amount varies from pair to pair: ISOGG's summary of the published data describes sibling pairs spread roughly between 37% and 62% around the average. So "half" is a center, not a guarantee.
ISOGG describes the result in terms of identical regions: siblings are half-identical where they inherited the same DNA from one parent, and fully identical where they inherited the same DNA from both. Regions where they inherited different DNA from both parents are the places where their ancestry labels can differ most.
From shared segments to different percentages
Ancestry companies look at your segments and compare them with reference panels (see how ancestry percentages actually work). If you and your sister received different segments from your father's mother, the regions assigned to those segments will differ, and so will your percentages.
The effect is smallest for large, widespread ancestries and largest for small contributions. Each generation back, an ancestor's expected share halves: about 12.5% from a great-grandparent, about 6.25% from a great-great-grandparent, and so on. The further back, the more likely a given ancestor left only small segments, or none that can be detected, in a particular child. One sibling may carry a trace region that the other does not.
A simple scenario shows the idea. Suppose one parent's parents came from two different regions. One child may happen to receive more segments from the first grandparent and another more from the second. Both are drawing from the same two sources, yet their estimates land in different places.
The pattern that explains the biggest gaps
Every child gets exactly half of their DNA from each parent, but how much that half varies depends on the parent's own ancestry:
- If one parent's own parents came from the same region, that parent's contribution looks much the same in every child.
- If another parent's parents came from different regions, each child's half is a different mixture of the two. This is where large sibling differences arise.
So when siblings differ a lot, look to the more mixed parent's side. This is a logical consequence of how inheritance works, rather than a special case.
What is the same in every sibling
- Mitochondrial DNA comes from the mother, so all siblings have her mtDNA haplogroup.
- The Y chromosome comes from the father, so brothers share his Y haplogroup, and sisters have none.
- Parents and grandparents are the same, so the broad story of the family is the same even when the numbers differ.
See haplogroups explained for these two lines.
What the match lists show
Cousin-match lists show the same variation. One genetic genealogist described sharing 95 centimorgans with a cousin match while his sister shared none with that same person, and a brother shared 68. Full siblings are not genetic copies: two of them can inherit different segments from a common grandparent, and so differ in which cousins they match. If you want the fullest picture of your ancestors, testing more than one sibling captures segments that one of you did not inherit.
When the gap is surprising
A very different result is usually just recombination. Occasionally it points to something else: the two people may be half-siblings who share only one parent, or a company update may have changed the reference data. Test results can reveal family information that no one expected, which we explore in how one relative's test affects the whole family. Take any surprise slowly, compare the match lists instead of the percentages, and consider talking to a genetic counselor if it is distressing.
How to compare results sensibly
- Compare the same company and reference version, not one sibling at one company and one at another.
- Focus on large regions and treat trace amounts as hints, not facts.
- Look at the shared DNA with the sibling if the company provides it. Roughly half is typical.
- Check for updates. Both results can change as the company improves its reference panel.
- Use paper records too. DNA and documents answer different questions.
If you are choosing a test for the whole family, our ancestry test comparison is a good starting point.
Bottom line
Siblings carry different samples of the same family DNA. Their ancestry reports differ because the samples differ, not because a lab made a mistake. Compare big regions, look for the mixed-ancestry parent behind large gaps, and treat small percentages as approximate.
Common questions
Is it normal for siblings to have different ancestry results?
Yes. Each child inherits a different random mix of each parent's DNA, so estimates built from those segments differ, especially for smaller regions.
How much DNA do full siblings share?
On average about half, according to ISOGG, but pairs vary around that average. The published range for sibling pairs is roughly 37% to 62% in the data ISOGG summarizes.
Could my sibling and I be half-siblings?
It is possible when the shared DNA is far below the typical range, but ancestry percentages alone cannot show it. Look at shared-DNA totals and talk to a genetic counselor if the result worries you.
Do identical twins get the same ancestry results?
They should get nearly identical results, because identical twins share almost the same DNA. Small differences can still arise from testing and processing.
Which sibling's result is right?
Both are correct estimates of the DNA each person inherited. They simply describe different samples of the family's ancestry.
Sources and verification
- ISOGG Wiki: Autosomal DNA statistics (expected sibling sharing about 50%; range reported by Visscher et al., 2006, roughly 37% to 62%)
- The Genetic Genealogist: leveraging siblings and cousins (example of siblings sharing different amounts with the same match)
The sibling sharing figures follow ISOGG's summary of published studies, not our own testing. The generational shares (12.5%, 6.25%) are expected averages and individual inheritance varies. The explanation about a mixed-ancestry parent is a direct consequence of how inheritance works and is not drawn from a specific study.
Medical disclaimer. GenomeTesting.org publishes educational content about genetics and consumer genomic testing. Nothing here is medical advice, a diagnosis, or a substitute for care from a qualified clinician or genetic counsellor. Decisions about testing, treatment, or interpreting your own results should be made with a healthcare professional who can see your full history.