A chart that relies on colour alone to distinguish its data excludes a significant share of its readers. Around one in twelve men and one in two hundred women have some form of colour-vision deficiency, most commonly difficulty distinguishing red from green. A chart with a red series and a green series, distinguished only by colour, is for those readers a chart with two identical series — the comparison the chart exists to show is simply absent. Accessible charting is not a decorative extra. It is the difference between a chart that communicates and one that does not, for a large minority of every audience.
Chart Builder offers accessible colour palettes for its bar, line, pie, scatter, and histogram charts, with PNG and SVG export. This post explains what makes a palette accessible, why labels matter more than legends, and the small design habits that make a chart readable by everyone — including readers who print in greyscale.
Why red-green fails, and what survives
The most common colour-vision deficiency confuses reds and greens, and to a lesser extent confuses some blues and purples. Palettes built on red-versus-green, or on subtle shades of similar hues, collapse for these readers: distinct series become indistinguishable, and a "good versus bad" colour coding becomes meaningless. The classic trap is a sales-versus-target chart in red and green, or a positive-versus-negative chart in the same pair — the exact distinction the chart is trying to emphasise is the one that disappears.
Palettes that survive use differences beyond hue: lightness and darkness, blue versus orange (a pair that most colour-blind readers distinguish well), and combinations that differ in more than one dimension at once. A well-designed accessible palette does not merely avoid red-green; it ensures that any two series differ in lightness as well as hue, so that even a fully greyscale rendering keeps them apart. That greyscale test is a useful habit: if your chart still reads when printed in black and white, its colour choices are doing their job rather than carrying the whole message.
Labels matter more than legends
Even with a good palette, colour should never be the only encoding. Direct labels — the series name written next to its line or on its bar, the value printed on the bar itself — make a chart readable regardless of colour perception, screen quality, or print fidelity. A legend, by contrast, forces the reader to match a colour swatch to a series and hold that mapping in mind while reading the chart — a task that fails the moment two colours look alike.
Direct labelling also helps every reader, not only colour-blind ones. A line chart with labels at the end of each line is faster to read than the same chart with a legend box, because the eye does not travel back and forth. A bar chart with values printed on the bars answers "how much" without estimation. A pie chart with labelled slices and percentages (for the few-category cases where a pie is appropriate) beats a pie with a colour-key legend. The principle is general: put the information where the reader is already looking, not in a separate key they must consult.
Patterns and shapes are the third encoding, useful where labels alone are crowded. Different line styles (solid, dashed, dotted) distinguish lines without colour; different bar patterns or point shapes distinguish series in dense charts. Use them sparingly — a chart with five patterns is as confusing as a chart with five similar colours — but for two or three series, a shape or pattern difference is cheap insurance.
Worked example: a fictional sales chart, before and after
A fictional company, Example Office Supplies, charts quarterly revenue for three product lines: Paper, Ink, and Furniture. The first version uses a red line for Paper, a green line for Ink, and a blue line for Furniture, with a legend box and no direct labels. For a reader with red-green deficiency, Paper and Ink merge into one muddy series, and the chart's central comparison — Paper overtaking Ink in the third quarter — is invisible.
The revised version changes three things. First, the palette switches to an accessible set: a dark blue, a warm orange, and a teal that differ in lightness as well as hue. Second, each line is labelled directly at its right-hand end with the product name and its final-quarter value. Third, the lines gain distinct styles: Paper solid, Ink dashed, Furniture dotted. The result reads in colour, in greyscale, and for a colour-blind reader — and it also reads faster for everyone else, because the labels sit where the eye already is.
The same habits apply to bars: an accessible palette, values printed on the bars, and series ordered meaningfully (largest to smallest, or in time order) rather than left to chance. And the same test applies before publishing: view the chart in greyscale, or with a colour-blindness simulator, and confirm every series is still distinguishable and every label still legible. If it fails that test, the palette or the labelling — not the reader — is the problem.
Honest limitations
An accessible palette improves readability; it does not guarantee it. Individual vision varies, screens and printers shift colours, and a palette that works for the most common deficiencies may still fail for rarer ones — which is why direct labels and patterns matter as much as colour choice. Accessibility also extends beyond colour: chart text must be large enough to read, contrast must be sufficient for low-vision readers, and the data behind a chart should be available as a table for readers who use screen readers or who need exact values. A chart is one presentation of data, not the only one, and the underlying numbers should never be locked inside an image alone.
For preparing data before charting, the workflow is clean first with CSV Cleaning Studio, then chart — and for resizing exports for the web, Image Optimizer can help. The background guide is how to choose the right chart.
FAQs
What makes a chart accessible to colour-blind readers?
Palettes that avoid red-green distinctions, differ in lightness as well as hue, and survive greyscale printing — combined with direct labels and, where useful, distinct line styles or patterns, so colour is never the only way to tell series apart.
Why do labels matter more than legends?
Because a legend forces the reader to match colours to series from memory, which fails when colours look alike. Direct labels put the series name and value where the reader is already looking, so the chart reads regardless of colour perception.
Should I test my chart in greyscale?
Yes. If every series is still distinguishable and every label still legible in greyscale, the chart's encoding does not depend on colour alone — a strong sign it will also read for colour-blind readers and in black-and-white print.
Does an accessible palette guarantee everyone can read my chart?
No. Vision varies, and rarer deficiencies may still affect a given palette. Combine an accessible palette with direct labels, patterns, sufficient text size, and a data table alongside the chart for exact values and screen-reader access.
human review pending client designation