Short answer
Make everything else quiet — usually grey — and give the one series or bar that matters a single saturated colour that also differs from the grey in lightness. A highlight works by contrast with its surroundings, so it fails if several colours compete, if the accent only differs in hue (which colour-vision deficiency and greyscale remove), or if the grey context is so pale it drops below readable contrast.
The eye goes to whatever differs most from its neighbours. In a chart with ten equally vivid series nothing stands out, so the reader has to consult the legend to find the one the text is about. Put nine of them in grey and one in a strong colour and the reader finds it before reading a word. This is the most effective use of colour in explanatory charts, and it inverts the categorical-palette problem: instead of making every colour equally distinct, you deliberately make one colour far more distinct than the rest. It also sidesteps the limit on how many categories a palette can hold, because only one or two need a real colour.
An orange line among grey lines looks obviously different to most readers, but much of that difference is hue and chroma, which greyscale printing takes away and colour-vision deficiency reduces. The table measures common accent choices against a mid-light grey context (#BBBBBB). Orange and sky blue differ strongly in colour, yet sit only 5–6 L* units from the grey — a contrast of about 1.2:1 — so in a mono print they merge with the context. Dark blue, vermilion and pure red keep a large difference in every condition because they are 20–30 L* units darker than the context; the same red against a darker grey near its own lightness would lose that advantage. The rule of thumb that follows: pick an accent at least 20–30 L* units away from the context grey, and prefer the darker side on a white background so the highlight also wins on contrast.
Greying out everything else is only acceptable if the grey marks remain usable. WCAG's non-text contrast criterion asks for 3:1 against adjacent colours for graphical objects needed to understand the content; a very pale grey that de-emphasises well may fail that. Whether it must pass depends on whether the context series are needed to understand the chart. If the chart's claim is 'this line rose faster than the others', the others are needed, and a grey around #949494 or darker on white keeps them at about 3:1 while still receding behind a strong accent. If the others are genuinely decorative, labels or a note should say so.
Two highlight colours can work when the story is a comparison (this line versus that one); three rarely do. More than that and the chart is a categorical chart with a grey background. Keep the highlight colour consistent across a report so readers learn that 'blue means the thing we are talking about', and avoid using a colour that already has a semantic meaning elsewhere — red for the highlight in a report where red also means 'loss' creates a false association. Finally, pair the colour with a label placed directly on the highlighted series; the label, not the colour, is what tells a screen-reader user or a greyscale reader which line matters.
| Accent | ΔE00 typical | ΔE00 deuteranopia (simulated) | Difference in L* | WCAG contrast with the grey |
|---|---|---|---|---|
| Vermilion #D55E00 | 33.3 | 30.0 | 21.7 | 2.01:1 |
| Blue #0072B2 | 33.7 | 36.1 | 29.9 | 2.70:1 |
| Orange #E69F00 | 28.4 | 28.1 | 5.3 | 1.17:1 |
| Sky blue #56B4E9 | 21.1 | 22.5 | 6.1 | 1.20:1 |
| Pure red #FF0000 | 36.4 | 29.5 | 22.6 | 2.08:1 |
Why: The accent differs from the grey context in hue but not in lightness.
Fix: Choose a darker accent, or thicken the highlighted line.
Why: Several series are coloured, so the accent has competitors.
Fix: Grey out everything except the one or two series the text discusses.
Why: The context grey is under 3:1 against the background.
Fix: Use a darker grey (about #949494 or darker on white) and rely on the accent's colour and weight for emphasis.
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 1.4.11 requires graphical objects that are needed to understand content to have a contrast ratio of at least 3:1 against adjacent colours.
Source: Web Content Accessibility Guidelines (WCAG) 2.2; Understanding Success Criterion 1.4.11: Non-text Contrast (WCAG 2.2)
Colourwise analysisModerate evidence
Against a #BBBBBB grey context, the Okabe–Ito orange and sky blue differ by only about 5–6 CIELAB L* units (WCAG contrast about 1.2:1) and would merge with it in greyscale, while blue #0072B2, vermilion #D55E00 and pure red stay 20–30 L* units darker and remain distinct under simulated deuteranopia and in greyscale.
Based on: Computed by Colourwise with the Viénot–Brettel–Mollon simulation, CIEDE2000 and CIELAB L*; see the table.
Caveat: Results depend on the particular context grey; a darker grey changes which accents work.
Source: Digital video colourmaps for checking the legibility of displays by dichromats; CIE (International Commission on Illumination) publications
Colourwise interpretationLimited evidence
Using one saturated accent against grey context series is an effective way to direct attention in explanatory charts, because salience depends on how much an element differs from its surroundings.
Based on: Colourwise's reading of widespread practice in data journalism and of basic visual-search findings that a single odd-one-out in colour is found quickly; no single controlled study is cited.
Caveat: The benefit depends on the accent differing in lightness as well as hue.
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.