Home DNA tests do work, but how well depends entirely on what you’re using them for. The raw technology behind these kits is highly accurate, with genotyping chips reading your DNA at above 99.5% accuracy. Where things get murkier is in the interpretation layer: what companies do with that data to estimate your ancestry, health risks, or traits. Some results are remarkably reliable, others are closer to educated guesses, and a few can be outright misleading.
How Home DNA Tests Read Your Genes
Every major consumer DNA company uses the same basic technology: a genotyping chip that scans specific locations across your genome. Instead of reading all 3 billion letters of your DNA (which is what whole genome sequencing does), these chips check hundreds of thousands of pre-selected spots where people commonly differ from one another. Think of it as reading every 10th word in a book rather than the whole thing. You get less information, but it’s dramatically cheaper and, for many purposes, enough to work with.
The chips themselves are precise. Reproducibility rates sit above 99.9%, meaning if you ran the same sample twice, you’d get nearly identical raw data. The accuracy of reading each individual spot is above 99.5%. So the physical process of extracting and reading your DNA from a saliva sample is not the weak link. The problems start when companies take that raw data and try to tell you something meaningful about your ancestry or health.
Ancestry Results Are Estimates, Not Facts
Ancestry breakdowns are the most popular feature of home DNA kits, and also the most misunderstood. When a test tells you you’re 40% Irish and 25% West African, those numbers aren’t measured directly from your DNA. They’re statistical estimates produced by comparing your genetic patterns against a reference database of people whose ancestry is already known.
This is where things get subjective. The results depend heavily on who’s in that reference database. If a company doesn’t have many reference samples from a particular region, it will struggle to identify ancestry from that area and may lump it into a neighboring category. Different companies use different reference populations, different statistical methods, and different ways of defining regional boundaries, which is why the same person can get noticeably different ancestry percentages from two different brands. Neither result is wrong exactly. They’re just different interpretations of the same data.
There’s a deeper issue, too. The genetic markers companies use to distinguish between populations are selected by researchers who have to decide in advance which groups count as distinct. As one analysis put it, these markers reflect the choices scientists made about who should be considered European or African, Greek or Nigerian, when they designed the test. The categories feel objective because they come with precise percentages, but they’re built on human decisions about how to carve up genetic variation that is, in reality, continuous and overlapping. Broad continental ancestry (European vs. East Asian vs. Sub-Saharan African) is quite reliable. The finer you go, the fuzzier the estimates become.
Health Risk Reports Vary Widely in Reliability
Health-related results from home DNA tests range from genuinely useful to dangerously misleading, depending on the specific condition and how the test was designed.
The FDA does review and clear certain categories of consumer genetic tests, including genetic health risk tests, pharmacogenetic tests (which predict how you might respond to certain medications), and cancer predisposition tests. These cleared tests meet specific standards for accuracy. However, ancestry tests, general wellness tests, and many low-risk results are not reviewed by the FDA before they go to market.
Even among FDA-cleared tests, the scope matters. The most well-known example is BRCA testing for breast and ovarian cancer risk. One major consumer test checks for only three specific BRCA mutations most common in people of Ashkenazi Jewish descent. For those three mutations, the false positive rate is very low, around 0.6%. But a retrospective study found that for BRCA mutations outside those three, the false positive rate from consumer testing was between 69% and 89.7%. That means the vast majority of people who got a positive result for other BRCA variants were told they had a mutation they didn’t actually carry. For mutations in other cancer-related genes, about 38% of positive results were false positives.
These numbers highlight a critical point: a positive health result from a consumer DNA test should never be treated as a diagnosis. The National Human Genome Research Institute recommends that any actionable finding from a consumer test be confirmed through clinical-grade genetic testing, which is held to stricter laboratory standards. Without that confirmation step, you could end up making medical decisions based on a result that doesn’t hold up.
Polygenic Risk Scores Are Especially Rough
Many consumer tests now offer risk scores for common conditions like heart disease, type 2 diabetes, or depression. These are polygenic risk scores, meaning they add up the small effects of hundreds or thousands of genetic variants to estimate your overall risk. The science behind them is real, but the practical value for any individual is limited.
These scores are calculated by weighting each risk variant based on findings from large genetic studies. The problem is that those studies have been conducted overwhelmingly in people of European descent, so the scores perform best for that group and can be significantly less accurate for people of other backgrounds. Differences in genetic patterns across populations mean the same score can mean different things depending on your ancestry. Two companies analyzing the same raw data can also produce different risk percentages simply because they made different methodological choices about how to build and calibrate the score.
A polygenic risk score can give you a general sense of where you fall on a spectrum, but it’s not a prediction. Your actual risk for most common diseases is shaped far more by lifestyle, environment, and family history than by any number a consumer test generates.
Paternity Tests: Accurate but Not Always Legal
For determining biological relationships, DNA testing is extremely reliable. Paternity tests typically report a “probability of paternity” of either 0% (not the father) or 99.9% (almost certainly the father). The underlying science is straightforward and well established.
The catch with home paternity kits is legal admissibility. Results from an at-home test are generally not accepted in court because there’s no verified chain of custody proving who actually provided the samples. If you need a paternity result for child support, custody, or any legal proceeding, the test must be completed at an approved facility where identities are verified and the process is documented. Home kits can satisfy personal curiosity, but they won’t hold up in a courtroom.
Third-Party Interpretation Tools Add Risk
Most consumer DNA companies let you download your raw genetic data, and a growing number of third-party websites offer to analyze it further, providing additional health reports, carrier screening, or trait predictions. These tools carry real risks.
The databases these services draw from may be outdated or may flag gene variants as disease-causing when they’re actually benign. One analysis found that third-party interpretation of raw data leads to a significant false positive rate for disease-causing BRCA variations specifically because harmless variants were being misidentified as dangerous. The result has been an increase in unnecessary medical appointments from people alarmed by inaccurate findings.
There are also privacy concerns. Uploading your data to public or semi-public databases makes it available to law enforcement, and many third-party services have vague policies about how your data will be used. A study of consumer genetic testing companies found that 67% provided insufficient information about how customer data would be used. Some share data with pharmaceutical companies while being vague about who receives it and how much they’re paid. In the U.S., these companies are currently under no obligation to notify you when they change their terms and conditions.
What You Can Trust and What You Can’t
The technology behind home DNA tests is sound. Your raw genotype data is read with high accuracy, and for straightforward questions like paternity or broad continental ancestry, the results are dependable. Where confidence should drop is in the layers of interpretation: fine-grained ancestry percentages are estimates that shift depending on the company, health risk reports for single-gene conditions carry meaningful false positive rates that demand clinical confirmation, and polygenic risk scores for common diseases are useful for population-level research but blunt instruments for individual prediction.
If you’re using a home DNA test for curiosity about your heritage, it will give you a reasonable sketch, not a photograph. If you’re using it for health information, treat any concerning result as a starting point for a conversation with a genetics professional, not an answer. And before you upload your raw data anywhere, read the privacy policy carefully, because once your genetic information is out there, you can’t take it back.