Short answer
Only with care. Red–green is the pair most likely to collapse for the roughly 1 in 12 men with a red–green colour-vision deficiency, its meanings are not universal (in some financial markets red means a rise), and the two often have similar lightness. Keep the familiar meanings if your audience expects them, but make them survive without hue: shift to blue versus red or orange, add signs, arrows or labels, and use colours that differ in lightness.
Red for loss and green for gain feel natural to many Western readers because traffic lights, accounting and user interfaces teach them. They are conventions, not perception. Apple's guidelines warn that red communicates danger in some cultures but has positive connotations in others, and in several East Asian stock markets red conventionally marks a price rise. Party colours, sports teams and brand palettes add further meanings that readers bring to a chart. Semantic colour is powerful when it matches the reader's expectation and actively misleading when it does not, so the first decision is whose conventions apply: the publication's audience, not the designer's.
The table computes pairs used for good/bad. The CSS named red and green, and the red and green of Material-style palettes, are strongly different for typical colour vision but fall to small differences under simulated deuteranopia, and their WCAG contrast with each other is close to 1:1 — they are almost the same lightness. A reader with the most common red–green deficiency therefore sees two similar mid-tones and has to guess which is which. Okabe and Ito's alternative — vermilion against bluish green — improves the separation under simulation, but the two are still almost the same lightness (about 1.1:1), so they merge in greyscale. Blue against orange keeps a large difference in every simulated condition and in lightness, because it varies along the blue–yellow axis that red–green deficiency leaves intact.
The practical compromise is to keep the colour a reader expects while making sure it is not the only carrier of meaning. Signed numbers (+4.2%, −1.8%), up and down arrows, 'gain' and 'loss' labels, or placing negative bars below an axis all communicate direction on their own; colour then reinforces rather than decides. WCAG 2.2 criterion 1.4.1 requires exactly this for web content: colour must not be the only visual means of conveying information. It is worth separating two uses that often get merged: status colours (on track, at risk, failed), where the categories are few and a legend or icon can do the work, and signed quantities, where a diverging scale and signed labels are the better tool.
Once a colour carries meaning, every other use of that colour in the same report inherits it. If red means 'below target' in the summary table, a red series in a later chart will be read as bad even if it is just one of several categories. Reserve the semantic colours: keep them out of categorical palettes, choose categorical colours that are clearly different from them, and write the mapping down in the report's style notes. When a chart has to show many categories and also flag good or bad, carry the good/bad with a second channel — marker shape, an outline or an icon — rather than trying to find more colours.
| Pair | ΔE00 typical | ΔE00 deuteranopia | ΔE00 protanopia | WCAG contrast between the two |
|---|---|---|---|---|
| CSS named red / green (#FF0000, #008000) | 72.2 | 15.0 | 10.6 | 1.28:1 |
| Material-style red / green (#D32F2F, #388E3C) | 66.0 | 5.5 | 19.5 | 1.20:1 |
| Okabe–Ito vermilion / bluish green (#D55E00, #009E73) | 54.4 | 20.1 | 18.5 | 1.13:1 |
| Okabe–Ito blue / orange (#0072B2, #E69F00) | 55.3 | 68.8 | 62.1 | 2.30:1 |
| Blue / red (#0072B2, #D55E00) | 49.6 | 62.5 | 55.2 | 1.34:1 |
Why: Red means 'up' in their convention.
Fix: Use signed numbers and arrows, and choose colours with your audience's convention in mind.
Why: The pair differs in hue only; lightness is nearly the same.
Fix: Add status icons or text, and choose colours that also differ in lightness.
Why: It was coloured red in a report where red carries a meaning.
Fix: Reserve semantic colours and keep them out of categorical palettes.
Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.
StandardStrong evidence
WCAG 2.2 success criterion 1.4.1 requires that colour is not the only visual means of conveying information, indicating an action, prompting a response or distinguishing a visual element.
Source: Web Content Accessibility Guidelines (WCAG) 2.2; Understanding Success Criterion 1.4.1: Use of Color (WCAG 2.2)
ConventionModerate evidence
Apple's Human Interface Guidelines advise using colour consistently for status, not using the same colour to mean different things, and considering that colours such as red carry different connotations in different cultures.
FactModerate evidence
Okabe and Ito report that about one in twelve Caucasian men, one in twenty Asian men and one in twenty-five African men have a red–green colour-vision deficiency.
Caveat: Prevalence estimates vary by study and population; these are the authors' summary figures.
Source: Color Universal Design (CUD) — How to make figures and presentations that are friendly to colorblind people; Colour Blindness / Colour Vision Deficiency (patient guidance)
Colourwise analysisModerate evidence
Common red/green status pairs have a WCAG contrast of little more than 1:1 with each other and a small CIEDE2000 difference under simulated deuteranopia, whereas blue paired with orange or vermilion stays well separated.
Based on: Computed by Colourwise with the Viénot–Brettel–Mollon simulation, CIEDE2000 and the WCAG contrast formula; see the table.
Caveat: Simulation represents full dichromacy; milder deficiencies see more difference.
Source: Digital video colourmaps for checking the legibility of displays by dichromats; Web Content Accessibility Guidelines (WCAG) 2.2
Reviewed 29 September 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.