How Ancestry Percentages Actually Work

Ancestry percentages feel like they should be simple math — half from each parent, quarters from each grandparent, and so on down the family tree. The reality is messier, and more interesting, because of how DNA actually gets passed down and how testing companies estimate origin from it.

Why Siblings Get Different Results

You inherit exactly half your autosomal DNA from each parent, but which specific half is randomized during a process called recombination, where chromosomes from your two parents shuffle and swap segments before being passed to you. Full siblings each receive a different random shuffle, so while you share roughly half your DNA with a sibling on average, the specific ancestral segments you each inherited from a given great-great-grandparent can differ meaningfully. Combined with the fact that more distant ancestors contribute smaller and smaller DNA segments that can effectively "wash out" over generations, it's entirely normal — not an error — for siblings to show noticeably different regional percentages, especially for smaller, more distant ancestral contributions.

Reference Panels: The Real Engine Behind the Percentages

Ancestry estimates work by comparing your DNA against reference panels — large collections of DNA samples from people with well-documented, multi-generational roots in specific regions. Your genome gets broken into segments, and each segment is statistically compared against these reference populations to estimate the most likely geographic origin. This means your result isn't a direct readout of "your ancestors came from X" — it's a probabilistic best match against whichever populations the company happened to include, and how they happened to define regional boundaries.

Why Your Results Can Change Over Time

It's common for ancestry percentages to shift after a company update, even though your DNA obviously hasn't changed. This happens because companies periodically expand their reference panels (adding more reference samples improves resolution, especially for historically underrepresented regions) and refine their statistical models. A result that said "broadly Northwestern European" in an earlier version of a reference panel might resolve into more specific sub-regional detail once the panel grows — reflecting an improvement in the underlying science, not a correction of an earlier mistake.

What Ancestry Percentages Are — and Aren't — Good For

• Good for: broad geographic and population history context, connecting with genetic relatives, satisfying genealogical curiosity, understanding population-specific disease risk patterns relevant to your background.
• Not good for: precise proof of a specific ancestor's origin, legal or citizenship documentation, or as a substitute for genealogical paper-trail research — DNA and documented family history answer related but different questions.

Whole genome sequencing has an advantage here worth noting: because it captures your complete genome rather than a fixed set of SNP markers, your raw data can be reanalyzed against improved reference panels in the future without retesting — the same reanalysis principle that applies to health-related variants applies to ancestry estimates too.

For more on getting the most from your raw data, see our guide to analyzing your raw genome data and our international buyer's guide if you're testing from outside the US.