Eye color statistics: we traced the “2% green eyes” figure, and nobody ever measured it
Eye color statistics audit: 0 of 20 global-percentage pages traced to a primary world measurement. See what the real data show.
You have probably read that green eyes belong to just 2% of the world. So have millions of other people, because nearly every eye color article repeats it.
The straight answer is that brown is the most common eye color worldwide and green is uncommon — but the precise global percentages repeated online have no traceable measurement behind them.
We checked 20 pages and followed their citations. None ended in a primary measurement of the global population. That makes the numbers untraceable, not automatically false: the direction is plausible, while the precision is unsupported.
Key numbers
- 0 of 20 citation chains in our audit reached a primary measurement of global eye-color shares (method below).
- 9 of 20 pages traced back to the uncited American Academy of Ophthalmology page “Eye Color: Unique as a Fingerprint.” AAO page
- 55% in 2017 versus 70–80% now is WorldAtlas’s changing brown-eye figure; neither version is sourced. 2017 archive · Current table
- 45% brown, 27% blue, 18% hazel, 9% green appeared in a 2014 US adult self-report poll. AAO summary of Harris Poll
- 57.4% versus 33.8% blue eyes were recorded among two birth cohorts of US non-Hispanic whites. Grant and Lauderdale
- 63.0% of 192 Stanford newborns had brown eyes at birth. NEST study
- 0.063%, roughly 1 in 1,587 was the measured prevalence of complete heterochromia in a US Military Academy portrait study. Dabkowski et al.
Table of contents
- What is the rarest eye color?
- How rare are green eyes, really?
- What is the most common eye color in the world?
- Where do the famous eye color percentages come from?
- What do real eye color measurements show?
- Why are eyes brown, blue, green or hazel?
- Can your eye color change?
- How rare is heterochromia?
- How to cite this page
- The bottom line
- Sources
What is the rarest eye color?
Among ordinary eye colors, green is usually named as the rarest. There is no traceable global measurement proving a precise green share. The AAO says “about 2% of the world’s population” without a citation, and that wording was already present in July 2019. AAO’s page
The AAO says albinism often produces light blue eyes; rare cases show blood vessels through the iris, making eyes look pink or red. Albinism is a medical condition, not a novelty shade. AAO’s eye color page
The useful answer is “green is uncommon,” not “green is exactly 2%.” Our most-attractive-eye-color guide goes deeper on that distinction.
The AAO may have an internal basis; the public chain does not show it.
How rare are green eyes, really?
Green eyes are uncommon, but their global percentage is unknown. The repeated figure is about 2%, while the pages we audited stated green shares of 2% to 3%, none backed by a primary measurement. Audit source node
In the Dutch Rotterdam Study, “green/brown spots” accounted for 18.98%, 18.89% and 31.71% in three cohorts, while the researchers also used a broad “intermediate” category. Categories changed, showing why denominator and classifier matter. Rotterdam Study
The 2014 Harris Poll gives 9% green among US adults who self-reported, alongside 18% hazel. That is context, not a world percentage: people classify hazel, green and light-brown irises differently, and an online poll is not an examination. Real World Appeal offers a free first-impression read before payment; it is not a clinical or scientific instrument and does not turn eye color into a score.
Directionally plausible; not measured.
What is the most common eye color in the world?
Brown is the most frequent eye color worldwide. MedlinePlus Genetics avoids a numeric worldwide share, and the geography review by Katsara and Nothnagel identifies data scarcity as a limitation, especially outside Europe. MedlinePlus Genetics · Katsara and Nothnagel
Consumer pages have repeated brown figures from “more than half” to 70–80%, including 79%, but our audit found no traceable primary measurement of the world population. WorldAtlas gave brown as 55% in an archived 2017 table and 70–80% in its current table, with no source at either date. WorldAtlas archive · WorldAtlas current table
“Worldwide” hides the denominator: the data are thinnest outside Europe. A European genetic study, a US poll and a Dutch cohort can all be real without adding up to a world census. The eye-shapes and attractiveness guide treats appearance categories as context-dependent, not universal laws.
Brown being most common is supported; the exact global percentage is not.
Where do the famous eye color percentages come from?
The familiar table is a citation cascade, not a global survey. In our audit of 20 pages, 15 cited a source, only 9 tied a source to the number, and 0 reached a primary global measurement.
| Figure you’ll see | Where the trail ends | Measured? |
|---|---|---|
| 2% green | Uncited AAO statement; some pages also use an uncited expert estimate | No global measurement found |
| 79% brown | Dorgaleleh et al. 2020 review, then a 2024 Scientific Reports introduction | No; review repetition |
| 70–80% brown | WorldAtlas current table, with no source | No source shown |
| 8–10% blue and 5% hazel | WorldAtlas, All About Vision and repeated lists | No global measurement found |
| Heterochromia 1% / under 200,000 Americans | WebMD and IFLScience claims; the 200,000 threshold is from rare-disease law | No matching count found |
“79% brown” moved from a 2020 iris-biochemistry review into a Scientific Reports introduction citing it, then into news and consumer pages. Newsweek’s “2% green” came from a DNA-matching dating-platform founder, not a study. Dorgaleleh et al. · Scientific Reports · Newsweek
The world’s eye-color percentages were never measured — they were copied. That criticizes the citation chain, not the claims that brown is common or green unusual. The AAO may have evidence; readers cannot see it in the public citation chain.
A paywalled source may hide an earlier reference, so “no traceable measurement” is fairer than calling the figures fabricated.
What do real eye color measurements show?
Real measurements exist, each with a defined population, method and category system. The table keeps denominators separate; none should be blended into a world percentage.
| Population measured | Eye color shares | How it was measured |
|---|---|---|
| US non-Hispanic whites born 1899–1905 vs 1936–51 | Blue: 57.4% vs 33.8% | NHANES government health surveys; recording method not stated in the abstract |
| Dutch older adults, Rotterdam Study RS1 | Blue 68.01%; intermediate 9.76%; brown 22.23% | Ophthalmologist classification; digital full-eye photographs also quantified |
| 23andMe European-ancestry customers | Blue 30.3%; intermediate 38.3%; roughly 31% brown remainder | Self-categorised seven-group color scale; >97% European ancestry |
| US adults, 2014 Harris Poll | Brown 45%; blue 27%; hazel 18%; green 9%; other 1% | Weighted self-report; more than 2,000 US adults |
| Stanford newborns | Brown 63.0%; blue 20.8%; green/hazel 5.7%; indeterminate 9.9%; partial heterochromia 0.5% | Iris color at birth in 192 healthy full-term newborns |
Grant and Lauderdale attributed the decline to assortative mating by ancestry, calling it the likely explanation. Grant and Lauderdale
Even ophthalmologist categories overlapped when researchers measured hue and saturation, and another study grouped blue, gray and green together as “light.” Liu et al. · Mengel-From et al.
Why are eyes brown, blue, green or hazel?
Eye color comes mainly from melanin in the front layers of the iris, combined with the way light scatters through the tissue. Brown eyes have more melanin; blue eyes have much less and contain no blue pigment. The color categories are a continuum, not four sealed biological boxes. MedlinePlus Genetics · AAO on blue eyes
The AAO explains green and hazel as light brown pigment interacting with scattered blue light. Common variants identified so far explain 53.2% of eye-color variation in Europeans, and the 2021 study found 124 independent associations in 61 genomic regions, 50 of them new. Simcoe et al.
A 2008 Copenhagen press release says the blue-eye mutation arose “6-10,000 years ago,” but that date is not in the study abstract. A 2025 review citing ancient-DNA work places the allele in northern Italy and the Caucasus 13,000-14,000 years ago. The date is disputed, not settled. Copenhagen press release · Bressan review
The biology is better established than the global percentages, but biological categories become less neat when researchers measure the iris continuously.
Can your eye color change?
Eye color often settles early, but “never changes after childhood” is too absolute. In the Louisville Twin Study, most people had stable eye color by age 6, while a 10% to 15% subpopulation of white participants continued changing through adulthood.
That study followed 1,513 twins using 15 painted glass-eye standards. Among 1,359 white twins, 3.8% to 8.6% shifted by two or more steps over three- to nine-year intervals between age 6 and early adulthood. This shows population-level change, not that every adult change is harmless. Bito et al.
In the Stanford cohort, brown was most common at birth, at 63.0%, contradicting the line that most babies are born blue. The AAO pages disagree about timing: one says color may darken over the first three years, while another says most babies have their lifelong color by about nine months. Stanford NEST study · AAO timeline
Wikipedia reports 17% changing by adulthood, but that figure cannot be confirmed from the Bito abstract; the full paper is paywalled. A new or sudden adult change is a clinician’s question, not a statistics question.
Some change after childhood is supported; no universal adult-change rate or explanation for every case is established.
How rare is heterochromia?
Complete heterochromia is rare in the measured studies: about 0.063%, or roughly 1 in 1,587 people. That estimate comes from 7 cases among 11,111 US Military Academy cadet portraits, and an earlier Vienna survey reported an almost identical figure.
The cadet estimate covers complete heterochromia in healthy young adults, not sectoral or central heterochromia. The researchers call prevalence poorly studied; an NIH StatPearls chapter says no prevalence is widely reported. Dabkowski et al. · StatPearls
The popular “1%” and “fewer than 200,000 Americans” claims do not resolve the problem. WebMD prints both figures even though they do not describe the same scale, while IFLScience links the 200,000 wording to a CMAJ teaching case whose abstract contains no prevalence figure. The US Food and Drug Administration uses fewer than 200,000 people as the Orphan Drug Act’s rare-disease threshold; treating that legal threshold as a headcount is an inference, not a demonstrated source trail. WebMD · FDA
Do not diagnose a different-looking eye from a screen; a clinician can assess a real change or symptom.
The 0.063% figure is the best measured estimate found here, not a census of every heterochromia form.
How to cite this page
Suggested citation: Real World Appeal. “Eye color statistics: we traced the ‘2% green eyes’ figure, and nobody ever measured it.” realworldappeal.com, September 2026. https://realworldappeal.com/blog/eye-color-statistics
The primary sources are linked throughout so readers can cite the underlying studies directly. The data and citation audit were retrieved and run on 13 September 2026.
The bottom line
Brown is the most common eye color worldwide, and green is uncommon. But a precise global table — 2% green, 79% brown, 8–10% blue and 5% hazel — should be labelled repeated estimates, not a measurement. The real data are narrower: a US poll, a Dutch cohort, European-ancestry customers, US birth cohorts and newborns answer different questions.
The world’s eye-color percentages were never measured — they were copied. Try Real World Appeal’s free first-impression read; results are visible before payment, with the caveat that it is not a clinical or scientific instrument.
Sources
- American Academy of Ophthalmology (2017). “Eye Color: Unique as a Fingerprint.” Linked source
- American Academy of Ophthalmology (2023). “Your Blue Eyes Aren’t Really Blue.” Linked source
- American Academy of Ophthalmology (2023). “Why Are My Eyes Changing Color?” Linked source
- Bito LZ, Matheny A, Cruickshanks KJ, Nondahl DM, Carino OB (1997). “Eye color changes past early childhood. The Louisville Twin Study.” Arch Ophthalmol 115(5):659-663. Linked source
- Bressan P (2025). “Why humans evolved blue eyes.” Front Psychol 16:1442500. Linked source
- Dabkowski M, Case J, Kloo I, Pickett J (2022). “Estimating the prevalence of heterochromia iridum from high-resolution digital yearbook portraits.” J Optom 15(3):248-250. Linked source
- Dorgaleleh S, Naghipoor K, Barahouie A, Dastaviz F, Oladnabi M (2020). “Molecular and biochemical mechanisms of human iris color: A comprehensive review.” J Cell Physiol 235(12):8972-8982. Linked source
- Ebrahimi, Gharravi, Aliyari, Emamian et al. (2024). “The association between iris color and refractive errors in children.” Sci Rep 14:2192. Linked source
- Grant MD, Lauderdale DS (2002). “Cohort effects in a genetically determined trait: eye colour among US whites.” Ann Hum Biol 29(6):657-666. Linked source
- Katsara MA, Nothnagel M (2019). “True colors: A literature review on the spatial distribution of eye and hair pigmentation.” Forensic Sci Int Genet 39:109-118. Linked source
- Liu F, Wollstein A, Hysi PG et al. (2010). “Digital quantification of human eye color highlights genetic association of three new loci.” PLoS Genet 6(5):e1000934. Linked source
- Ludwig CA, Callaway NF, Fredrick DR, Blumenkranz MS, Moshfeghi DM (2016). “What colour are newborns’ eyes? Prevalence of iris colour in the Newborn Eye Screening Test (NEST) study.” Acta Ophthalmol 94(5):485-488. Linked source
- MedlinePlus Genetics, U.S. National Library of Medicine (2026). “Is eye color determined by genetics?” Linked source
- Mengel-From J, Børsting C, Sanchez JJ, Eiberg H, Morling N (2010). “Human eye colour and HERC2, OCA2 and MATP.” Forensic Sci Int Genet 4(5):323-328. Linked source
- Simcoe M, Valdes A, Liu F et al. (2021). “Genome-wide association study in almost 195,000 individuals identifies 50 previously unidentified genetic loci for eye color.” Sci Adv 7(11):eabd1239. Linked source
- U.S. Food and Drug Administration (2026). “Rare Diseases at FDA.” Linked source
- University of Copenhagen (2008). “Blue-eyed humans have a single, common ancestor.” EurekAlert! news release 747660. Linked source
- WebMD (2026). “What Is Heterochromia?” Linked source
- WorldAtlas (2017). “Which Eye Color is the Most Common in the World?” Linked source
- WorldAtlas (2026). “The World’s Population By Eye Color.” Linked source
Frequently asked questions
What is the rarest eye color?
Green is commonly named as the rarest ordinary eye color, but no traceable measurement establishes a global percentage. Red or pink-looking eyes are associated with rare albinism, a medical condition; see our eye-color attractiveness guide.
What percentage of people have green eyes?
The familiar claim is about 2% worldwide, but our audit found no primary measurement behind it. A 2014 US self-report found 9% green, which is a US figure rather than a world estimate; the attractiveness myths fact-check explains why denominators matter.
What is the most common eye color?
Brown is the most frequent eye color worldwide according to MedlinePlus Genetics, but the exact worldwide share has not been measured. A 2014 US poll found 45% brown among adults who self-reported their color; our first-impression guide puts rarity in its proper context.
Can your eye color change as an adult?
Usually eye color is stable by age 6, but a 10% to 15% subpopulation of white participants in the Louisville Twin Study kept changing through adolescence and adulthood. Changes in adulthood can also have medical significance, so a clinician is the right person to ask about a new change; our eye-contact guide covers the social side, not diagnosis.
How rare is heterochromia?
Complete heterochromia was measured at about 0.063%, roughly 1 in 1,587, in a study of US Military Academy cadet portraits. That estimate does not measure partial or central heterochromia; our photo first-impression read is about expression and presentation, not a medical assessment.