Color Blindness Statistics: What 1 in 12 Actually Means for You

In a room of 30 people there is a 72.8% chance at least one has a colour vision deficiency. If that room is all men, it is 91.8%. Nine men is already a coin flip.

The Headline Numbers

The commonly cited prevalence figures, and the ones used throughout this site, are:

About 1 in 12 men — roughly 8% — have some form of colour vision deficiency.
About 1 in 200 women — roughly 0.5%.
Around 300 million people worldwide.
Roughly 1 in 30,000 have complete colour blindness, or achromatopsia.
Around 1 in 10,000 have a tritan (blue-yellow) deficiency, affecting men and women about equally.

Where these come from, and how firm they are. These are widely repeated estimates drawn from population studies of predominantly European-ancestry groups, and prevalence genuinely varies between populations — several studies report lower rates in African and East Asian populations. Treat 8% as a reasonable planning figure for a general Western audience, not a universal constant. Everything computed below is exact arithmetic on that assumption; if your audience differs, the inputs change and so do the outputs.

The Number That Actually Matters

“8% of men” is hard to act on. What changes a decision is the answer to: is there someone in this room? That is a different calculation, and it climbs much faster than people expect.

For a group of n people where each has probability p, the chance that at least one is affected is 1 − (1 − p)ⁿ. Using 8% for men, 0.5% for women, and a 50/50 split giving 4.25% per person:

Probability at least one person in a group has a colour vision deficiency
Group	Size	Mixed	All men
A small standup	5	19.5%	34.1%
A product team	10	35.2%	56.6%
A department	20	58.0%	81.1%
A classroom	30	72.8%	91.8%
A company all-hands	50	88.6%	98.5%
A conference track	100	98.7%	100.0%

Three thresholds worth remembering. In a mixed group it becomes more likely than not at 16 people. In an all-male group — still the reality of plenty of engineering teams — it happens at nine. And a mixed group hits 95% at 69 people.

So the honest reading of “8% of men” is not “a small minority”. It is: if you have ever presented a red-green chart to a room of thirty, someone in that room probably could not read it, and almost certainly did not tell you.

Turning It Into Traffic

The same arithmetic applied to visitor numbers, at the 4.25% blended rate:

Expected number of colour blind visitors by traffic level
Monthly visitors	Affected, on average
100	4
1,000	43
10,000	425
100,000	4,250
1,000,000	42,500

A modest site with 10,000 monthly visitors serves roughly 425 people who cannot rely on your colour coding. That is not an edge case — for most teams it is a larger group than the entire cohort they run usability tests with.

It is worth saying what this number is not. It is an expectation, not a headcount, and it assumes your audience resembles the population the base rates came from. A site whose users skew heavily male — much of gaming, engineering tooling, motorsport — should use a figure closer to 8% and roughly double every row.

Why the Split Between Men and Women Is So Wide

The genes for the red and green photopigments sit on the X chromosome. Men have one X, so a single altered copy produces the deficiency. Women have two, and a working copy on either is generally enough — which is why the rate drops from about 1 in 12 to about 1 in 200, and why it commonly passes from grandfather to grandson through an unaffected mother.

Blue-yellow deficiency is the exception. It is carried on chromosome 7, which is not sex-linked, so it affects men and women at similar rates — our genetics guide covers the inheritance patterns in more detail.

The Statistic Nobody Publishes

Prevalence tells you how many people are affected. It says nothing about how often your specific design fails them, which is the number you can actually change.

That one is measurable. When we ran every pair of Tailwind CSS's 500-weight colours through simulation, 69 of the 136 pairs came back confusable — and four were identical. Your palette has a failure rate, and unlike the population figures it is one you can compute exactly and then fix.

The useful move. Stop arguing about whether 8% justifies the work, and measure your own colours instead. One failing pair in a status indicator affects every single one of those users, every time they load the page.

Frequently Asked Questions

How many people are color blind?

Around 300 million worldwide is the commonly cited figure, based on roughly 8% of men and 0.5% of women. These are estimates from population studies of mainly European-ancestry groups, and prevalence varies between populations, so treat them as planning figures rather than precise counts.

What percentage of my users are color blind?

For a general audience with a roughly even gender split, about 4.25% — so around 425 people per 10,000 visitors. If your audience skews male, as it does in gaming, engineering tooling and motorsport, use a figure closer to 8% and roughly double that.

How likely is it that someone on my team is color blind?

In a mixed group of 30, there is a 72.8% chance at least one person is. In an all-male group of 30 it is 91.8%, and it passes 50% at just nine men. The probability climbs far faster with group size than the raw percentage suggests.

Why are men so much more likely to be color blind?

The red and green photopigment genes sit on the X chromosome. Men have one X, so a single altered copy causes the deficiency; women have two, and a working copy on either is usually enough. Blue-yellow deficiency is carried on chromosome 7 instead, so it affects both sexes about equally.

Is color blindness becoming more or less common?

There is no strong evidence of a meaningful shift. It is inherited rather than environmental, so prevalence changes only slowly with population genetics. What has changed is detection — screening in schools and for certain jobs means more people know about it than a few generations ago.

Do these statistics justify redesigning my product?

That is the wrong framing, and it is a question you can replace with a measurement. Rather than arguing about whether a percentage clears some bar, test your actual colours: a palette either has confusable pairs or it does not, and fixing one is usually a change of shade rather than a redesign.