Research & testimony
DNA Test Errors
Paternity tests are not infallible. Below is one father's account, followed by documented public cases of wrong results.
My story — Daniel Joseph McKelvie
Me and my daughter, Conceptual Radiance High Day
A DNA test told me I was not my daughter's father. I believe that result was a false negative.
She was born 9 months and 2 days after our conception — the natural length of a pregnancy. The other man's time with her mother was about 3 weeks earlier. For him to be the father, she would have been born roughly 3 weeks late, and she was not.
The test result contradicts the dates, from conception to her first birthday.
Personal account of the site owner.
Documented cases of false results
Canadian lab knowingly sold wrong prenatal paternity results (2024)
A CBC News investigation found Viaguard Accu-Metrics' prenatal tests routinely identified the wrong biological fathers; the owner said on hidden camera the results were "never that accurate."
CBC News99.9% result later corrected to 0% (New York, 2025)
DNA Diagnostics Center issued a corrected report months after a first paternity result. The woman, who acted on the first result, is suing.
NBC NewsExcluded twice, matched by another lab (Oklahoma Supreme Court, 2011)
LabCorp twice reported a man was not the father; a different laboratory later matched him. The court allowed the negligence claim to proceed.
Berman v. LabCorpDisputed prenatal test ended a relationship (Kentucky, 2020)
A 2014 prenatal test said the boyfriend was not the father; later tests contradicted it.
McDonald v. DNA Diagnostics CenterTwo mismatches were natural mutations
A commercial lab reported a negative result from two mismatches. A retest showed probability of paternity above 99.9999%; the first lab reported incorrectly.
Forensic Science International: GeneticsFather excluded by repeat tests — then confirmed (chimerism)
A man carrying two distinct genomes was repeatedly excluded before further testing affirmed his paternity.
Journal of Assisted Reproduction and GeneticsPaternity pseudo-exclusion at 6 markers (2023)
A father excluded at six DNA markers was proven the father once other tissue samples were tested.
European Journal of Medical GeneticsLydia Fairchild: DNA said she was not her children's mother (2002)
She was accused of fraud until chimerism explained the mismatch.
Embryo Project EncyclopediaDNA chips miss the mark on rare variants (BMJ, 2021)
For very rare variants, only 16% of consumer-style SNP chip positives were confirmed by sequencing.
BMJUp to 40% of consumer DNA raw-data findings may be false
Reported clinical re-testing could not confirm many variants flagged in consumer test data.
MIT Technology ReviewFalse DNA test led a father to reject his daughter
A UK report of a family harmed by a wrong result.
The Telegraph
The genotyping chip error
Consumer DNA companies read hundreds of thousands of DNA points using SNP "chips". Customers can download this raw data and upload it to other online matching and analysis sites.
A 2021 BMJ study of nearly 50,000 people found chips were over 99% accurate for common variants, but for very rare variants only 16% of positive calls were real — and 20 of 21 consumer-test customers had at least one false "disease" variant in their data.
Clinical re-testing reported in 2018 could not confirm about 40% of variants flagged in consumer raw data. Once wrong raw data is uploaded and shared, the error can travel with it.
Standard court paternity tests use a different method (STR markers), so chip errors and paternity errors are separate problems — but both show DNA results can be wrong.
Known causes of wrong paternity results
- Sample mix-ups and laboratory reporting errors
- Natural mutations read as mismatches
- Chimerism — one person carrying two sets of DNA
- Close male relatives producing misleading matches
- Unsupervised at-home sample collection