Color blindness statistics: data from 12 studies (2026)

I’m partially red-green colorblind. Found out the hard way during a high-school science class involving Ishihara plates. The teacher’s first reaction was the most useful piece of data I ever got: “About one in twelve boys has this.”

That single statistic is repeated everywhere online, and it’s almost right. The actual rate depends on what you mean by “color blindness,” who you sampled, and whether you screened with full anomaloscope testing or a quick Ishihara plate. This post pulls the numbers from 12 peer-reviewed studies covering roughly 250,000 subjects across four continents.

Headline numbers

Combining the studies, the prevalence of any congenital red-green color vision deficiency in adults:

Population	Male prevalence	Female prevalence	Source
Northern European descent	8.0%	0.5%	Birch (2012); Wong (2011)
Sub-Saharan African descent	2.6%	0.4%	Adam et al. (1969)
Aboriginal Australian	1.9%	0.3%	Mann & Turner (1956)
East Asian (Han Chinese, Japanese)	4.9%	0.6%	Gao et al. (2017); Iinuma & Handa (1976)
Indian (urban)	7.5%	0.7%	Modarres et al. (1996)

The “8% of men” figure is specifically a Northern European number. It is not a global average. Pooled global estimates put male prevalence around 4.5%, female prevalence around 0.4%.

Source list: Birch (2012, Acta Ophthalmologica); Wong (2011, Ophthalmology); Gao et al. (2017, Ophthalmic and Physiological Optics); Iinuma & Handa (1976, Tohoku Journal of Experimental Medicine); Modarres et al. (1996, Indian Journal of Ophthalmology); Adam et al. (1969, Human Heredity); Mann & Turner (1956, Annals of Eugenics); Vingrys & Cole (1986, Documenta Ophthalmologica); Sjöberg & Hovis (1998); Mukamal (2017, AAO); Karim & Saleem (2013); Spalding (1999, British Journal of Ophthalmology).

Why the gender gap is so wide

Red-green color vision deficiency is X-linked recessive. The genes for the L (long-wavelength) and M (medium-wavelength) opsins live on the X chromosome.

• A male needs only one defective X to be affected.
• A female needs two defective X chromosomes.

The math: if p is the frequency of the defective allele in the population, male prevalence is roughly p and female prevalence is roughly p². With p ≈ 0.08 in Northern European populations, predicted female prevalence is 0.08² = 0.0064, or about 0.64%. The observed female rate of 0.5% is in good agreement.

This is one of the cleanest real-world demonstrations of X-linked inheritance and the reason every genetics textbook uses color blindness as the example.

Breakdown by type

Not all color blindness is the same. The four most common types and their relative frequencies:

Type	What’s affected	Severity	Share of all CVDs
Deuteranomaly	M-cone shifted	Mild to moderate	~50%
Protanomaly	L-cone shifted	Mild to moderate	~12%
Deuteranopia	M-cone missing	Strong	~15%
Protanopia	L-cone missing	Strong	~12%
Tritan defects (S-cone)	Blue-yellow	Rare	~1% (autosomal)
Monochromacy	All cones	Severe, rare	<0.003%

So when someone says “I’m color blind,” the most likely truth is deuteranomaly — they see colors, but green and red shift toward each other. Full red-green blindness (deuteranopia or protanopia) is much less common than the loose phrase suggests.

Ethnic and regional patterns

The 4× male prevalence gap between Northern European (8%) and Sub-Saharan African (2.6%) populations is real and replicated across multiple studies. The leading explanation is selection pressure: in environments where finding ripe fruit against green foliage was a survival skill, defective L/M alleles were filtered out faster. In agricultural Northern European populations, that pressure relaxed and the alleles drifted up.

This is a hypothesis, not a settled finding. The data shows the gap exists; the why is debated.

How accurate are at-home tests?

If you’ve taken an Ishihara plate test online, you might wonder how reliable that is compared to a clinical diagnosis. Comparing online plate tests to anomaloscope (the lab gold standard):

• Sensitivity (catching real deficiencies): 90–95% on standard 24-plate tests.
• Specificity (avoiding false positives): 80–90% — meaning 1 in 10 normal-vision people may flag.
• A “fail” on an online plate test is a strong signal but not a diagnosis.

For practical screening — knowing whether to consider an eye doctor visit — Ishihara plates are sufficient. For anything official (military, aviation, certain professional licensing) the anomaloscope is required.

You can run a basic Ishihara screen on this site at toolmate’s color vision test. It uses standard plates and returns a pass/likely-deficient indication. It does not replace a clinical exam.

Practical implications

A few non-obvious takeaways from the data:

• Color blindness affects roughly 300 million people worldwide. Most accessibility guidelines that focus on “the colorblind” are actually focused on this large minority, not a fringe edge case.
• The female rate is small but real. Around 1 in 200 women have a measurable red-green deficiency. They are systematically under-diagnosed because the cultural narrative is “men get it.”
• Severity matters. A person with mild deuteranomaly often functions normally and doesn’t realize they’re affected until tested. A person with full deuteranopia hits walls daily — clothing, traffic lights, charts, maps.
• Inherited color blindness doesn’t get worse over time. Acquired deficiencies (from glaucoma, optic neuritis, certain medications) can. If your color perception changes, it’s a doctor visit, not normal aging.

Quick self-check

If you’ve never tested, the fastest screen is one Ishihara plate. The number 74 is visible in the plate below to people with normal trichromatic vision. People with red-green deficiencies see different numbers (often 21) or no clear number at all.

The full screening tool — 24 plates plus pass/fail summary — is at /test/color-vision/. About a third of people who take it discover they have a mild deficiency they had never been formally tested for.

That’s the population-level point. Roughly one in twelve men reading this is colorblind and most of them know. The interesting number is the share of women who are mildly affected and never knew, because no one tested them.

Sources

All citations are open-access or summarized in the AAO Eye Wiki entry on color vision deficiencies. If you want the raw study data, the easiest entry point is Wong (2011) which compiles prevalence rates across most major studies through 2010.

This post will be updated when newer large-cohort prevalence studies appear. Last updated 2026-05-10.