Full siblings share approximately 50% of their DNA on average. That number comes with a surprisingly wide range, though. Two siblings from the same parents can share anywhere from roughly 32% to 54% of their DNA, depending on which segments each child happened to inherit. This variation explains why siblings can look strikingly different from one another despite having the exact same parents.
Why 50% Is an Average, Not a Guarantee
Each parent has two copies of every chromosome, and each child gets one copy from mom and one from dad. The copy a child receives is essentially random. Two siblings might inherit the same copy of a chromosome from their father, or they might each get a different one. Multiply that coin-flip across all 23 chromosome pairs from both parents, and the actual overlap between any two siblings varies quite a bit.
At any given spot in the genome, siblings fall into one of three categories: they inherited the same chromosome copy from both parents (making that region identical), they share a copy from one parent but not the other (partially identical), or they received completely different copies from each parent (no shared DNA at that spot). Roughly one quarter of the genome ends up fully identical between siblings, while the rest is a mix of partially shared and unshared regions.
On top of that randomness, chromosomes don’t pass down as clean, intact copies. Before a parent’s egg or sperm cell forms, matching chromosomes swap segments with each other in a process called recombination. This shuffling creates unique hybrid chromosomes for each child, so even when two siblings inherit the “same” chromosome from a parent, the actual DNA sequence along that chromosome can differ. Recombination is the reason no two siblings (aside from identical twins) ever share exactly the same DNA.
Siblings Compared to Other Relatives
The 50% average for full siblings puts them in the same ballpark as a parent-child relationship, which is also 50%. The key difference is that parent-child sharing is fixed at 50% every time, because a child always gets exactly half from each parent. Sibling sharing fluctuates around that number because of the random assortment and recombination described above.
Half-siblings, who share one biological parent, average about 25% of their DNA. In practice, half-sibling sharing ranges from roughly 16% to 33%, measured in centimorgans (a unit geneticists use to quantify shared DNA segments). The average for half-siblings is around 1,759 centimorgans, with a range spanning 1,160 to 2,436 cM. For comparison, full siblings typically share around 2,600 to 3,400 cM.
Identical twins are the only siblings who share virtually 100% of their DNA. They form from a single fertilized egg that splits in two, so they start with the same genetic blueprint. Fraternal twins, on the other hand, are no more genetically similar than any other pair of full siblings. They just happened to share the womb at the same time.
How Sisters and Brothers Differ in DNA Sharing
The roughly 50% figure applies to the 22 pairs of non-sex chromosomes, but the sex chromosomes add an interesting wrinkle. Fathers have one X and one Y chromosome. Mothers have two X chromosomes. A father passes his X to every daughter and his Y to every son, with no recombination on most of the X (or Y) in those cases.
This means two sisters who share the same father will always carry an identical copy of his X chromosome. Their mother’s X contribution may differ between them, since female X chromosomes do recombine during egg cell formation. But the paternal X is handed down intact, giving sisters a guaranteed extra block of shared DNA beyond what brothers share.
Two brothers, by contrast, each receive their father’s Y chromosome, which is much smaller and carries far fewer genes than the X. Brother-sister pairs share neither a matching X nor a matching Y from their father. These sex-chromosome patterns don’t dramatically change the overall 50% average, but they do mean sisters tend to land slightly higher in shared DNA than brother pairs.
What DNA Tests Actually Show
Consumer DNA tests from companies like 23andMe and AncestryDNA measure shared DNA by scanning hundreds of thousands of genetic markers across the genome. They look for continuous stretches of matching markers, requiring a minimum segment length of about 7 centimorgans to count as a real match rather than a coincidence. Full siblings show approximately 50% shared DNA, and the test labels the match as “Siblings.” Half-siblings show roughly 25% and get labeled “Half-siblings.”
These predictions are generally reliable for close relationships but can occasionally be ambiguous. A pair of half-siblings at the high end of their sharing range can overlap with full siblings at the low end. When this happens, the testing company may present multiple possible relationships and let you sort it out with additional family context or testing.
When Siblings Share More or Less Than Expected
In rare cases, a genetic event called uniparental disomy can shift how much DNA siblings share. This occurs when a child inherits both copies of a particular chromosome from the same parent instead of one from each. It typically results from errors during egg or sperm cell formation, where an extra chromosome copy gets included and a “rescue” mechanism in the early embryo discards one copy to restore the normal count. If the wrong copy gets discarded, the child ends up with two copies from one parent.
Uniparental disomy for a single chromosome is uncommon and usually affects only a small fraction of total DNA. It can sometimes cause health issues if it leads to two copies of a faulty gene or disrupts genes that function differently depending on which parent they came from. But for sibling comparison purposes, it would only nudge the overall sharing percentage by a small amount on one chromosome.
Sibling DNA in Forensic Investigations
Law enforcement sometimes uses a technique called familial DNA searching, where a crime scene sample is compared against a DNA database to find not just direct matches but close relatives of a potential suspect. Full siblings and parent-child pairs produce the strongest signals in these searches. Half-sibling relationships can technically be detected but are often excluded from search protocols because the weaker genetic overlap makes them less reliable as leads.
A familial DNA search never serves as evidence on its own. It generates an investigative lead, pointing detectives toward a family rather than a specific person. Investigators then obtain a direct DNA sample from the actual suspect and run a standard one-to-one comparison. Only that exact match is used in court. The sibling or parent connection simply narrows the search, much like a witness description might narrow a suspect pool without proving guilt.