How this instrument works
Real eye color is a polygenic trait, not a simple one- or two-gene trait: the largest genome-wide association study to date, covering almost 195,000 people, identified 61 distinct genomic regions associated with eye color, and a model built from every common genetic variant known still explains only about 53% of the variation between people (Simcoe et al., Science Advances, 2021). That study's participants, like most large-scale eye-color genetics research to date, were overwhelmingly of European ancestry, so both those findings and the simplified blue/green/brown model below are least well studied — and least representative — for people of other ancestries, where brown-eye variation is the global majority phenotype and remains comparatively understudied at the genetic level. This calculator instead uses a deliberately simplified three-allele teaching model — brown dominant over green, green dominant over blue — purely to illustrate basic Mendelian inheritance. Every percentage below is an illustration of a teaching concept, not a genetic forecast for a real pregnancy.
The model assigns each parent one of three alleles per gene copy — brown, green, or blue — with brown masking both green and blue, and green masking blue. A blue-eyed parent has an unambiguous genotype (blue/blue), so the calculator reads their allele contribution directly. Brown-eyed and green-eyed parents are genuinely ambiguous: the observed eye color could come from more than one underlying genotype, so the calculator has to make an assumption about which genotype each one represents before it can work out what gets passed on.
For a brown-eyed parent, this calculator follows the same convention used elsewhere in eye-color calculators: it assumes a 50/50 split between the two heterozygous genotypes that produce brown eyes (brown paired with green, and brown paired with blue), deliberately leaving out the homozygous brown-brown possibility as a conservative simplification. For a green-eyed parent, only two genotypes can produce green eyes in the first place — homozygous green-green and heterozygous green-blue — so the calculator splits 50/50 between those two directly, with no third option to set aside.
Because of the reduced allele count and the 50/50 assumption for ambiguous phenotypes, read the percentages here as an educational approximation of how inheritance works at a basic level, not a precise prediction. This tool has no role in paternity testing, clinical genetic counseling, or any legal proceeding, and a newborn's eye color can keep changing for the first year or more of life regardless of what any calculator reports.
- Set Mother's eye color: Blue, Green, or Brown.
- Set Father's eye color: Blue, Green, or Brown.
- Review the transmitted-allele breakdown for each parent — how often the model has them passing on a brown, green, or blue allele.
- Read the three child probabilities: brown eyes, green eyes, and blue eyes, which always add to 100%.
- Change either parent's selection to compare scenarios, and re-read the disclaimers above before treating any figure as more than a teaching illustration.
Worked example — two brown-eyed parents
Both parents select Brown. Under this model a brown-eyed parent's genotype is treated as ambiguous, split 50/50 between brown-green and brown-blue (excluding homozygous brown-brown), so each parent transmits a brown allele 50% of the time (mBr = fBr = 0.50), a green allele 25% of the time (mGr = fGr = 0.25), and a blue allele the remaining 25% of the time (mBl = fBl = 0.25).
Child brown probability is 1 minus the chance neither parent passes brown: 1 − (1 − 0.50) × (1 − 0.50) = 1 − 0.25 = 0.75, or 75.0%. Child blue probability is the chance both parents pass blue: 0.25 × 0.25 = 0.0625, or 6.25%. Green takes what is left over: 100% − 75.0% − 6.25% = 18.75%.
This result — two brown-eyed parents having roughly a 1-in-16 chance of a blue-eyed child under the model — echoes a real, well-documented genetics phenomenon: brown-eyed parents who each carry a hidden recessive allele can have a blue-eyed child. The exact 6.25% figure, though, is a property of this simplified three-allele teaching model and its 50/50 convention, not a measured population statistic.
Questions
Is this calculator a scientifically accurate way to predict my baby's eye color?
No, and it is not trying to be. Eye color is polygenic — the largest genome-wide study to date (almost 195,000 people) found 61 distinct genomic regions associated with it, and even a model using every known common variant explains only about 53% of the variation between people (Simcoe et al., Science Advances, 2021). This calculator uses a simplified three-allele teaching model to illustrate basic Mendelian inheritance, so its percentages should be read as an educational illustration, not a scientific prediction of your own child's eye color.
How does the 50/50 assumption for ambiguous parents work?
It differs slightly by phenotype because the underlying genotype options differ. A brown-eyed parent could genuinely be one of three genotypes, so the calculator splits 50/50 between the two heterozygous ones (brown-green and brown-blue) and sets aside homozygous brown-brown as a conservative simplification. A green-eyed parent only has two possible genotypes in this model — homozygous green-green and heterozygous green-blue — so the 50/50 split is simply between those two, with nothing excluded.
Why can two brown-eyed parents have a blue-eyed baby?
Because a brown-eyed parent can still carry a hidden blue allele underneath their visible brown eyes. If both parents happen to carry that hidden allele and both pass it to the same child, the child can end up with two blue alleles and blue eyes even though neither parent shows blue. This calculator models that possibility, which is why two Brown selections still show a non-zero blue percentage rather than a flat 100% brown result.
Can I use this calculator for paternity testing or any legal purpose?
No. This tool is an educational illustration of a simplified three-allele Mendelian model, built to show basic inheritance patterns, not a diagnostic, forensic, or legal instrument. Real paternity determination relies on DNA profiling across many independent genetic markers, performed by an accredited laboratory, and has nothing in common with this page's simplified eye-color arithmetic.
Why do the three probabilities always add up to 100%?
Because brown, green, and blue are the only three outcomes this simplified model allows, and every child must land in exactly one of them under the model's own rules. Green is deliberately calculated as whatever remains after brown and blue are subtracted from 100%, which is also why the three figures always reconcile exactly rather than needing independent verification.
Does a newborn's eye color at birth predict their eye color as an adult?
Not reliably. Many babies, especially those with lighter skin tones, are born with eyes that appear blue or gray simply because melanin has not yet fully developed in the iris; the eventual adult color can take six months to three years to settle in, and it can shift somewhat further from there. This calculator's output describes a modeled adult phenotype, not what a newborn's eyes will look like on day one.
What if a parent's eyes are hazel or another color not listed?
This calculator only accepts Blue, Green, or Brown because those are the three phenotypes its simplified allele model was built around. Hazel and other intermediate shades reflect additional genetic and pigment factors this three-allele model does not represent, so there is no accurate way to map them onto one of the three listed options without misrepresenting the underlying genetics.
Where does this three-allele brown/green/blue model come from?
It is a long-standing classroom simplification of eye-color inheritance, built on a basic dominance order — brown over green, green over blue — that predates modern genome-wide research and is still commonly used to teach Mendelian concepts like dominant and recessive alleles. It is not itself a peer-reviewed genetic model; the peer-reviewed picture is the much more complex polygenic one described above and in the Simcoe et al. 2021 study.
References
- Simcoe et al. 2021, GWAS of ~195,000 people, 61 loci for eye color — Science Advances
- MedlinePlus Genetics (NIH) — Is eye color determined by genetics?
Read this first: This instrument computes a screening figure from population formulas — it is not a diagnosis, and it cannot see the whole picture a clinician can. Use it to inform a conversation, not to replace one.