An estimated 2.59% of children and adolescents worldwide have a congenital color vision deficiency, a figure that rises to 4.38% among males (Jeong et al., Ophthalmology 2025 meta-analysis). The interfaces they use are not keeping up: low-contrast text was detected on 83.9% of the top one million home pages in early 2026, up from 79.1% a year earlier (WebAIM Million 2026). Color is also a quiet failure point in science communication, where 85.6% of 66,253 eLife figures contained at least one potentially problematic color pair (Stevens et al., eLife 2024). With US public entities facing WCAG 2.1 AA deadlines in 2027 and 2028 and digital accessibility litigation already tracked in our web accessibility lawsuit statistics, color is a compliance issue as much as a design one. We aggregated data from the National Eye Institute, WHO, W3C WCAG 2.2, the WebAIM Million 2026, peer-reviewed meta-analyses in Ophthalmology and PLOS One, and population studies from Israel, China, Taiwan, Vietnam and Africa.
TL;DR
- Global congenital CVD prevalence among children and adolescents: 2.59% (Ophthalmology, Jeong et al. 2025)
- Male prevalence 4.38% versus 0.64% for females (Ophthalmology, Jeong et al. 2025)
- About 1 in 12 men have color vision deficiency (National Eye Institute)
- Deutan (green) defects are the most common type: 3.66% of males (Ophthalmology, Jeong et al. 2025)
- 3.06% of 916,388 Israeli military recruits had CVD (British Journal of Ophthalmology 2025)
- Only 19.2% of Nigerian students with CVD knew they had it (Healthcare review, Almustanyir et al. 2025)
- 34% of people with CVD in an Australian study said it influenced their career choice (Healthcare review, Almustanyir et al. 2025)
- 83.9% of top home pages have low-contrast text, the most common WCAG failure (WebAIM Million 2026)
- Low-contrast instances per home page rose 15% to 34 in one year (WebAIM Million 2026)
- WCAG AA text contrast minimum is 4.5:1; AAA is 7:1 (W3C WCAG 2.2)
- 12.8% of reviewed eLife figures were definitely problematic for deuteranopes (eLife, Stevens et al. 2024)
- 19.8% of low vision survey respondents reported color vision deficiency (WebAIM Low Vision Survey 2018)
1. Global Prevalence: How Many People See Color Differently
The newest global estimate is lower than the folk number most designers quote. The “8% of men” figure describes people of Northern European descent; a 2025 narrative review notes that red-green CVD affects up to 8% of males and 0.5% of females of Northern European descent, with lower rates in Asian and African populations (Almustanyir et al., Healthcare 2025). The pooled worldwide figure, built from 56 studies across 21 countries and 1,703,619 participants, lands at 2.59%.
For product teams the practical reading is the same either way: in any audience with a typical male share, several percent of users cannot rely on hue alone. Region matters, though. Oceania (4.37%) and Africa (2.86%) led regional estimates in the Ophthalmology meta-analysis, which complicates the old assumption that CVD is mainly a European issue.
| Metric | Value | Source |
|---|---|---|
| Global congenital CVD prevalence (children and adolescents) | 2.59% (95% CI 2.20-3.03) | Ophthalmology, Jeong et al. 2025 |
| Prevalence in males | 4.38% | Ophthalmology, Jeong et al. 2025 |
| Prevalence in females | 0.64% | Ophthalmology, Jeong et al. 2025 |
| Prevalence, European descent | 2.77% | Ophthalmology, Jeong et al. 2025 |
| Prevalence, African descent | 2.69% | Ophthalmology, Jeong et al. 2025 |
| Highest regional prevalence (Oceania) | 4.37% | Ophthalmology, Jeong et al. 2025 |
| Prevalence in Africa (region) | 2.86% | Ophthalmology, Jeong et al. 2025 |
| Men with color vision deficiency | About 1 in 12 | National Eye Institute, Color Blindness page |
Context: the National Eye Institute notes that men are at much higher risk, that people who are white are more likely to have CVD, and that acquired cases can follow retinal or optic nerve damage, brain tumors, radiation treatment or cataracts. CVD sits inside a much larger vision picture: WHO estimates at least 2.2 billion people have a near or distance vision impairment (WHO fact sheet, February 2026).
2. Prevalence by Country and Population
Country-level studies show why a single “global” number should never be the only input to a design decision. A university app in Taiwan and a defense recruitment portal in Israel serve populations whose male CVD rates differ by more than a percentage point, and the sex gap is consistently enormous. In Vietnam, male health science students had 16-fold higher odds of CVD than female students (OR = 16.60; 95% CI 5.32-82.8) (Scientific Reports 2026).
The largest single dataset here is the Israeli one: 916,388 recruits aged 16 to 18 screened between 2000 and 2020, of whom 28,001 had CVD (British Journal of Ophthalmology 2025). Prevalence was highest among recruits whose mothers were born in the former Soviet Union, a reminder that ancestry, not geography, drives inherited rates.
| Metric | Value | Source |
|---|---|---|
| CVD among Israeli military recruits (n = 916,388) | 3.06% (28,001 recruits) | British Journal of Ophthalmology, Barayev et al. 2025 |
| CVD among recruits whose mothers were born in the former Soviet Union | 4.98% | British Journal of Ophthalmology, Barayev et al. 2025 |
| Pooled CVD prevalence across Africa (16 studies) | 2.71% (95% CI 2.28-3.14) | PLOS One, Tilahun et al. 2024 |
| Africa pooled prevalence by sex | 2.13% males, 0.34% females | PLOS One, Tilahun et al. 2024 |
| Highest pooled national prevalence in Africa (Ethiopia) | 3.63% | PLOS One, Tilahun et al. 2024 |
| Chinese college students screened as CVD (n = 17,303) | 2.20% (4.11% males, 0.16% females) | International Journal of Ophthalmology, Gao et al. 2023 |
| Congenital red-green CVD in Southern Taiwan | 3.46% males, 0.14% females | International Journal of Molecular Sciences, Kuo et al. 2023 |
| Vietnamese health science students with confirmed CVD | 1.9% (28 of 1,491) | Scientific Reports 2026 |
The African meta-analysis is available in full at PLOS One. Note that the Chinese and Taiwanese studies are from 2023: most recent available data for those populations.
3. CVD Types, Severity and Linked Eye Conditions
Type matters more than headcount for palette choices. Deutan defects alone reach 3.66% of males, which is why red and green pairs are the classic failure: a status dot that flips from green to red can be invisible to the most common CVD group. Severity splits by sex too. Males show more dichromacy (missing one cone type entirely), while females more often have milder anomalous trichromacy (Ophthalmology 2025).
CVD rarely travels alone. The Israeli recruit data found higher rates of amblyopia, strabismus and ametropia among people with CVD (all p < 0.001), so a user who struggles with hue may also struggle with acuity or contrast. That is a strong argument for designing with luminance contrast, not just hue swaps.
| Metric | Value | Source |
|---|---|---|
| Deutan-type CVD prevalence, males | 3.66% | Ophthalmology, Jeong et al. 2025 |
| Deutan-type CVD prevalence, females | 0.46% | Ophthalmology, Jeong et al. 2025 |
| Dichromacy prevalence, males | 1.59% | Ophthalmology, Jeong et al. 2025 |
| Anomalous trichromacy prevalence, females | 0.38% | Ophthalmology, Jeong et al. 2025 |
| Red-green CVD in adults over 60 | 3.73% (protanomaly 1.51%, protanopia 1.76%, deuteranopia 0.45%) | International Journal of Ophthalmology, Hashemi et al. 2023 |
| Amblyopia, CVD vs normal color vision | 1.28% vs 0.71% | British Journal of Ophthalmology, Barayev et al. 2025 |
| Strabismus, CVD vs normal color vision | 1.09% vs 0.83% | British Journal of Ophthalmology, Barayev et al. 2025 |
| High ametropia, CVD vs normal color vision | 3.18% vs 2.30% | British Journal of Ophthalmology, Barayev et al. 2025 |
The older-adult figures come from Hashemi et al. 2023, most recent available data for that age group. Age adds a second layer: WCAG itself cites the loss of contrast sensitivity that accompanies aging as one reason for its contrast thresholds (see section 6).
4. Awareness, Careers and Quality of Life
The biggest gap is not prevalence but awareness. When 80.8% of Nigerian students with CVD did not know they had it, users cannot be expected to switch on a colorblind mode they do not know they need. Accessible defaults matter more than optional toggles. A 2026 narrative review of health professions education reached a similar conclusion, finding that few programs screen for CVD or offer accommodations (Imhoff et al., The Clinical Teacher 2026).
The consequences show up in careers and daily life. In an Australian study summarized by the Healthcare 2025 review, about a third of people with CVD said it shaped their career choice and almost a quarter were barred from specific occupations. In a Chinese university cohort, a majority reported a measurable quality-of-life impact on one of two instruments (Gao et al., International Journal of Ophthalmology 2023).
| Metric | Value | Source |
|---|---|---|
| Nigerian students with CVD aware of their condition | 19.2% aware, 80.8% unaware | Healthcare (Basel) review, Almustanyir et al. 2025 |
| Indian students who had heard of CVD | 75.2% (few could define it accurately) | Healthcare (Basel) review, Almustanyir et al. 2025 |
| People with CVD whose career choices were influenced (Australia) | 34% | Healthcare (Basel) review, Almustanyir et al. 2025 |
| People with CVD explicitly barred from specific occupations (Australia) | 24% | Healthcare (Basel) review, Almustanyir et al. 2025 |
| Chinese students with CVD whose quality of life was affected | 67.63% (CBQoL), 42.27% (utility analysis) | International Journal of Ophthalmology, Gao et al. 2023 |
| Odds of a good QoL score, CVD vs normal color vision (India) | OR = 0.31 | Indian Journal of Ophthalmology, Male et al. 2023 |
| Saudi dentists with red-green CVD | 6.0% | Frontiers in Dental Medicine, Almarzouki et al. 2026 |
| Saudi dentists with borderline color vision findings | 15.3% | Frontiers in Dental Medicine, Almarzouki et al. 2026 |
Correction devices are not a substitute for accessible design. A 2022 systematic review of 16 studies found EnChroma glasses, Chromagen filters and EnChroma Cx-14 do not provide clinically significant evidence of improved subjective color perception (Health Science Reports 2022). The dental figures come from Frontiers in Dental Medicine, where shade matching is part of daily clinical work.
5. Low-Contrast Text on the Web: WebAIM Million 2026
The web was slowly improving on contrast and then reversed. Low-contrast text fell from a 2021 peak of 86.4% to 79.1% in 2025, then jumped back to 83.9% of home pages in February 2026, according to the WebAIM Million 2026. It remains the single most common detected WCAG failure, ahead of missing alt text (53.1%) and missing form labels (51%).
Page complexity is part of the story. The average home page grew from 1,257 to 1,437 elements in a year, and every new badge, chip and gray caption is another chance to miss 4.5:1. Because low contrast hits CVD users, older users and anyone reading on a dim screen at once, it is the cheapest accessibility fix with the widest reach. Readers tracking the screen-side of the problem can compare these numbers with our digital eye strain statistics.
| Metric | Value | Source |
|---|---|---|
| Home pages with low-contrast text (2026) | 83.9% | WebAIM Million 2026 |
| Home pages with low-contrast text (2025) | 79.1% | WebAIM Million 2026 |
| Home pages with low-contrast text (2019) | 85.3% | WebAIM Million 2026 |
| Peak year for low-contrast text (2021) | 86.4% | WebAIM Million 2026 |
| Distinct low-contrast instances per home page | 34, up 15% from 2025 | WebAIM Million 2026 |
| Home pages with detected WCAG 2 failures | 95.9% (94.8% in 2025) | WebAIM Million 2026 |
| Average detected errors per home page | 56.1 (56,114,377 total) | WebAIM Million 2026 |
| Average elements per home page | 1,437 (1,257 in 2025) | WebAIM Million 2026 |
Context: WebAIM’s six most common error categories account for 96% of all detected errors, so contrast is part of a small, fixable cluster rather than a long tail. These figures measure text contrast only; they say nothing about charts, status dots or form errors that rely on hue alone (SC 1.4.1 Use of Color), so they are a floor for color-related barriers, not a ceiling.
6. The Contrast Rules: WCAG 2.2 Thresholds and Legal Deadlines
The thresholds are not arbitrary. W3C sets 4.5:1 for normal text because it compensates for the contrast loss of users with roughly 20/40 vision, which it describes as typical acuity of elders at roughly age 80, and explicitly includes congenital or acquired color deficiencies in that rationale (W3C Understanding 1.4.3). Because the ratio is computed on luminance, it holds regardless of which specific color deficiency a user has. Ratios range from 1:1 to 21:1.
Color-only meaning has its own rule. Success Criterion 1.4.1 Use of Color is Level A, the most basic tier, and the W3C guidance treats a 3:1 lightness difference as a valid way to distinguish inline links from surrounding text. Legally, the US Department of Justice’s ADA Title II rule makes WCAG 2.1 AA the technical standard for state and local government web content and mobile apps (ADA.gov).
| Metric | Value | Source |
|---|---|---|
| Minimum contrast, normal text (SC 1.4.3, AA) | 4.5:1 | W3C WCAG 2.2 Understanding 1.4.3 |
| Minimum contrast, large text (18pt, or 14pt bold) | 3:1 | W3C WCAG 2.2 Understanding 1.4.3 |
| Enhanced contrast, normal text (SC 1.4.6, AAA) | 7:1, sized for about 20/80 acuity | W3C WCAG 2.2 Understanding 1.4.6 |
| UI components and graphical objects (SC 1.4.11, AA) | 3:1 against adjacent colors (2.999:1 fails) | W3C WCAG 2.2 Understanding 1.4.11 |
| Inline link vs surrounding text, when color distinguishes links | 3:1 | W3C WCAG 2.2 Understanding 1.4.1 |
| Focus indicator (SC 2.4.13) | 3:1 change, area of at least a 2 CSS pixel perimeter | W3C What’s New in WCAG 2.2 |
| WCAG 2.2 publication | 5 October 2023, 9 new success criteria | W3C What’s New in WCAG 2.2 |
| ADA Title II compliance dates (WCAG 2.1 AA) | April 26, 2027 (50,000+ persons); April 26, 2028 (smaller and special districts) | US DOJ, ADA.gov Title II web rule |
The 1.4.11 strict-threshold note is in W3C Understanding 1.4.11, and the AAA rationale is in Understanding 1.4.6. Focus appearance details are in What’s New in WCAG 2.2.
7. Color-Safe Figures, Apps and Assistive Options
Charts and app palettes are where color carries the most meaning, and they fail at scale. Of 66,253 eLife images analyzed, only 2.6% were grayscale and 85.6% included at least one color pair with potentially problematic contrast; in a manually reviewed set, 12.8% were definitely problematic for deuteranopes (Stevens et al., eLife 2024). The encouraging part: distance between colors and labels mitigated problems in about half of cases, and the share of definitely problematic articles declined from 2012 to 2021. Labels and direct annotation do real work, the same principle behind tactile graphics for math and colorblind modes covered in our game accessibility features statistics.
Mobile apps are now getting automated color repair. DroidPalette fixed 66.60% of 497 real color-related issues across 105 Android apps (ASE 2025), and the earlier Iris tool reported a 91.38% repair success rate (arXiv 2308.09029). Users with low vision overrepresent CVD and lean on contrast controls: 51.4% of WebAIM low vision respondents used some high contrast mode, and 71.2% of those preferred light text on dark backgrounds (WebAIM Low Vision Survey #2).
| Metric | Value | Source |
|---|---|---|
| eLife images with at least one potentially problematic color pair | 85.6% (56,816 of 66,253) | eLife, Stevens et al. 2024 |
| Reviewed eLife images definitely problematic for deuteranopes | 12.8% (636 of 4,964) | eLife, Stevens et al. 2024 |
| PubMed Central test images definitely problematic | 5.2% (104 of 1,191) | eLife, Stevens et al. 2024 |
| Cases where color distance / labels mitigated problems | 54.2% / 48.4% | eLife, Stevens et al. 2024 |
| DroidPalette repair success on Android color issues | 66.60% (497 issues, 105 apps) | ASE 2025, DroidPalette |
| Iris repair success on Android color issues | 91.38% (9 of 40 pull requests merged) | arXiv 2308.09029, Iris 2023 |
| Low vision survey respondents with CVD | 19.8% (248 respondents) | WebAIM Survey of Users with Low Vision #2, 2018 |
| Low vision respondents using a high contrast mode | 51.4% | WebAIM Survey of Users with Low Vision #2, 2018 |
Outlier note: a 2026 notch filter trial reported improved color discrimination in 83% of 12 male subjects on the Farnsworth-Munsell 100 Hue test, but the authors call the results preliminary and performed no group-level statistical testing (Diagnostics 2026). The WebAIM low vision survey is from 2018: most recent available data.
Summary: Color Vision Deficiency and Accessible UI by the Numbers
| Metric | Value | Source |
|---|---|---|
| Global congenital CVD prevalence (children and adolescents) | 2.59% | Ophthalmology, Jeong et al. 2025 |
| CVD prevalence, males vs females | 4.38% vs 0.64% | Ophthalmology, Jeong et al. 2025 |
| Men with color vision deficiency | About 1 in 12 | National Eye Institute |
| Deutan-type CVD, males | 3.66% | Ophthalmology, Jeong et al. 2025 |
| People with near or distance vision impairment | At least 2.2 billion | WHO fact sheet 2026 |
| CVD among 916,388 Israeli recruits | 3.06% | British Journal of Ophthalmology 2025 |
| Pooled CVD prevalence in Africa | 2.71% | PLOS One, Tilahun et al. 2024 |
| Chinese college students with CVD | 2.20% | International Journal of Ophthalmology 2023 |
| Male vs female odds of CVD (Vietnam) | OR = 16.60 | Scientific Reports 2026 |
| Nigerian students with CVD unaware of it | 80.8% | Healthcare (Basel) review 2025 |
| People with CVD barred from some occupations (Australia) | 24% | Healthcare (Basel) review 2025 |
| Home pages with low-contrast text | 83.9% | WebAIM Million 2026 |
| Low-contrast instances per home page | 34 (up 15%) | WebAIM Million 2026 |
| Home pages with detected WCAG 2 failures | 95.9% | WebAIM Million 2026 |
| WCAG AA normal text contrast | 4.5:1 | W3C WCAG 2.2 |
| WCAG AAA normal text contrast | 7:1 | W3C WCAG 2.2 |
| UI component and graphic contrast | 3:1 | W3C WCAG 2.2 |
| eLife images with a potentially problematic color pair | 85.6% | eLife 2024 |
| DroidPalette Android color repair success | 66.60% | ASE 2025 |
| Low vision respondents with CVD | 19.8% | WebAIM Low Vision Survey 2018 |
Methodology and Sources
- Jeong YD et al., Global Prevalence of Congenital Color Vision Deficiency among Children and Adolescents, 1932-2022, Ophthalmology 2025: doi.org/10.1016/j.ophtha.2025.07.031
- Barayev E et al., Colour vision deficiency and amblyopia, strabismus and ametropia, British Journal of Ophthalmology 2025: doi.org/10.1136/bjo-2023-324564
- Tilahun MM et al., Prevalence of color vision deficiency in Africa, PLOS One 2024: doi.org/10.1371/journal.pone.0313819
- Almustanyir A et al., A Global Perspective of Color Vision Deficiency, Healthcare (Basel) 2025: doi.org/10.3390/healthcare13162031
- Gao JG et al., CVD among Chinese college students and quality of life, International Journal of Ophthalmology 2023: doi.org/10.18240/ijo.2023.09.23
- Hashemi H et al., Red-green CVD in adults above 60, International Journal of Ophthalmology 2023: doi.org/10.18240/ijo.2023.09.22
- Kuo HK et al., Congenital CVD in Southern Taiwan, International Journal of Molecular Sciences 2023: doi.org/10.3390/ijms242015247
- CVD among first-year health science students in Vietnam, Scientific Reports 2026: doi.org/10.1038/s41598-026-53435-1
- Male SR et al., CVD-QoL in an Indian population, Indian Journal of Ophthalmology 2023: doi.org/10.4103/ijo.ijo_1975_22
- Almarzouki M et al., CVD among Saudi dentists, Frontiers in Dental Medicine 2026: doi.org/10.3389/fdmed.2026.1906369
- Imhoff M et al., CVD in Health Professions Education, The Clinical Teacher 2026: doi.org/10.1111/tct.70402
- Male SR et al., Color vision devices: systematic review and meta-analysis, Health Science Reports 2022: doi.org/10.1002/hsr2.842
- Hsu JC et al., Effects of Notch Filters in Correcting CVD, Diagnostics 2026: doi.org/10.3390/diagnostics16091347
- Stevens HP et al., Identifying images in the biology literature that are problematic for people with a CVD, eLife 2024: elifesciences.org/articles/95524
- DroidPalette, Characterizing and Repairing Color-Related Accessibility Issues in Android Apps, ASE 2025: doi.org/10.1109/ase63991.2025.00169
- Iris, Automated and Context-Aware Repair of Color-Related Accessibility Issues for Android Apps, 2023: arxiv.org/abs/2308.09029
- National Eye Institute, Color Blindness: nei.nih.gov
- WHO, Blindness and vision impairment fact sheet (10 February 2026): who.int
- WebAIM, The WebAIM Million 2026: webaim.org/projects/million
- WebAIM, Survey of Users with Low Vision #2 (2018): webaim.org/projects/lowvisionsurvey2
- W3C, Understanding WCAG 2.2 SC 1.4.1, 1.4.3, 1.4.6 and 1.4.11, and What’s New in WCAG 2.2: w3.org/WAI
- US Department of Justice, ADA Title II web and mobile accessibility rule: ada.gov
- Data watch: the 2.59% global estimate covers children and adolescents in 56 studies (1932-2022) and should not be read as an all-age adult figure; the widely quoted 8% of men applies to Northern European descent. Awareness and career figures (Nigeria, India, Australia) are original studies as summarized in the Almustanyir 2025 review, not re-analyzed here. WebAIM Million results come from automated testing of home pages only and measure text contrast, not color-only meaning (SC 1.4.1). The Chinese, Taiwanese, over-60 (Hashemi) and Indian studies are from 2023 and the WebAIM low vision survey is from 2018 (most recent available data). ADA Title II compliance dates are as published on ADA.gov at the time of writing. Popular advocacy estimates of total colorblind people worldwide were excluded because their underlying methodology is not disclosed. Some publisher pages (ACM Digital Library) were not reachable; abstracts were read via Europe PMC, OpenAlex and arXiv.
Last updated: October 4, 2026. We update this roundup quarterly, and the next refresh is expected when WebAIM publishes the WebAIM Million 2027 report.