A chart with poor color choices doesn't just look bad to color blind users—it becomes completely unintelligible. Data series merge together, trends disappear, and your carefully crafted insights become noise. Color blindness friendly chart colors aren't a nice-to-have; they're essential for effective data communication.
Why Standard Chart Palettes Fail
The default palettes in Excel, Google Charts, and most visualization libraries weren't designed with color blindness in mind. They typically include several shades of red and green that become nearly identical for the 8% of men with red-green color blindness.
Even worse, some "accessible" palettes only consider one type of color blindness. A palette that works for deuteranopia might still fail for protanopia or tritanopia. True accessibility means testing across all major types.
Proven Color Blindness Friendly Palettes
IBM Design Colorblind Safe Palette
Designed and tested by IBM's accessibility team for use across all visualization types:
Works for protanopia, deuteranopia, tritanopia, and normal vision.
Wong Colorblind Safe Palette
Published in Nature Methods, designed specifically for scientific visualization:
8-color palette tested extensively in academic publishing.
Tol Qualitative Palette
Paul Tol's palette, optimized for maximum distinctiveness:
Excellent for charts with many data series.
Beyond Colors: Additional Accessibility Techniques
Even the best color blindness friendly chart colors should be supplemented with other visual cues. Here's how:
Patterns & Textures
Use different fill patterns (solid, striped, dotted, hatched) alongside colors. Each data series becomes distinguishable even in grayscale or for users with severe CVD.
Line Styles
For line charts, vary dash patterns (solid, dashed, dotted) and line thickness. Combined with color, this creates unmistakable differentiation.
Direct Labels
Place labels directly on chart elements rather than relying on color-matched legends. This eliminates the need to match colors entirely.
Data Point Shapes
Use different shapes for scatter plots and line charts: circles, squares, triangles, diamonds. Each shape is distinct regardless of color.
Chart-Specific Guidelines
Bar and Column Charts
- Limit to 4-6 colors when possible. More than that strains distinguishability for everyone.
- Use value labels on or near bars to reduce reliance on legend matching.
- Consider grayscale for simple comparisons—variations in brightness work for all users.
Pie Charts
- Direct labeling is essential. Don't make users match pie slice colors to a distant legend.
- Pull out key slices to add positional differentiation.
- Consider alternatives: Bar charts often communicate the same data more accessibly.
Line Charts
- Vary line styles (solid, dashed, dotted) in addition to colors.
- Use different marker shapes at data points.
- Label lines directly at the end or alongside the data.
Heat Maps and Gradients
- Use single-hue gradients (light to dark blue) rather than diverging color scales.
- Avoid red-green gradients entirely. Blue-yellow or purple-orange work better.
- Include value labels where space permits.
Testing Your Chart Palettes
Before publishing any data visualization, test it across all major CVD types:
- Upload a screenshot to our Color Blindness Simulator
- Check each CVD type: Protanopia, deuteranopia, tritanopia, and achromatopsia
- Verify distinguishability: Can you tell all data series apart in each simulation?
- Test the palette alone: Use the Palette Generator to verify your color choices before building charts
Test Your Data Visualizations
Upload your charts and see exactly how they appear to users with color vision deficiency.