Search for color psychology and you will find the same infographic reproduced a thousand times. Blue means trust. Green means growth. Red creates urgency and increases conversions by some confidently stated percentage. It is repeated so often that it feels like settled knowledge.
Most of it is not. Some of it is genuinely supported. This post tries to separate the two honestly, because designers make real decisions on the basis of these claims.
The famous button study did not say what people think
The single most-cited piece of evidence in this area is an A/B test in which a red button outperformed a green one. It gets summarised as "red converts better".
What the test actually showed is narrower. On that page, with that layout, a red button outperformed a green one — on a page where green was already the dominant color of the surrounding design. The red button was the only element that stood out. What was being measured was contrast against context, not an intrinsic property of red.
Run the same test on a red-heavy page and a green button would very plausibly win. The generalisable finding is not "use red". It is "make the primary action visually distinct from everything around it", which is a statement about salience, and which does replicate.
What the research reasonably supports
Salience drives attention. An element that differs from its surroundings gets looked at first. This is well established in visual attention research and it is the mechanism behind most apparent "color increases conversion" results. It is also why an accent color stops working the moment you use it everywhere.
Some associations are learned and real within a culture. In contexts where red has consistently meant error or stop, people process red warnings faster. This is not something inherent to the wavelength — it is a convention people have learned, and it works because they learned it. That distinction matters, because it means the association is only reliable within the population that shares it.
Color aids memory and recognition. There is reasonable evidence that colored images and materials are recognised somewhat better than grayscale equivalents. This supports the practical value of consistent brand color, though the effect sizes are modest.
Preference for color varies with context, not just hue. Research on color preference finds that people's liking for a color is heavily influenced by what they associate it with, and by saturation and lightness rather than hue alone. A "nice blue" and an "unpleasant blue" are both blue.
What does not hold up well
Fixed emotional meanings per hue. The chart mapping each color to an emotion does not survive scrutiny. Effects that appear in one study frequently fail to replicate, and reported effect sizes are typically small, context-dependent and sensitive to how the question was asked.
Cross-cultural universality. Color associations differ substantially across cultures. White signals purity in some contexts and mourning in others. Red indicates good fortune in parts of East Asia and danger in much of Europe — and in several Asian stock markets red indicates a price rise, the opposite of the Western convention. Any claim beginning "red universally means" is wrong somewhere important.
Precise conversion numbers. Claims like "orange increases signups by 32.5%" are almost always a single unreplicated test on one page, stripped of its context and repeated until it sounds like a law. Treat any specific percentage with no linked methodology as decoration.
Blue suppresses appetite. A perennial claim with weak support that gets recycled endlessly. If you have read that fast food chains use red and yellow to stimulate hunger — those chains chose colors that are highly visible from a road, decades ago, and kept them for brand continuity.
Why the myths persist
Three reasons, and recognising them makes the pattern easy to spot.
The claims are memorable and satisfying. "Blue means trust" is a complete, repeatable idea. "It depends on context, culture and salience" is correct and much harder to put on a slide.
Citation drift compounds. A modest finding gets summarised in a blog post, that post gets cited without its caveats, and after several rounds a tentative result has become a stated fact with no path back to the original study.
And confirmation is easy. Banks do use blue, often. But banks also chose blue partly because other banks used blue, and because it looks conservative in print. Institutional conformity is a simpler explanation than a psychological effect, and it fits the evidence at least as well.
What to do instead
None of this means color choice is arbitrary. It means the useful principles are different from the popular ones.
Design for salience, deliberately. Decide what should be noticed first and give it the strongest contrast on the page. Reserve your accent color for actions. This is the mechanism actually underlying most claimed color effects, and you can control it directly.
Follow conventions your users already know. Red for destructive actions and green for confirmation work not because of hue psychology but because people have learned them in software. Breaking a learned convention has a real cost. Following it is free.
Match the category, then differentiate within it. Users have expectations about what a finance product or a children's app looks like. Meeting the broad expectation aids recognition; standing out within it aids memorability. This is a positioning decision, not a psychological one.
Test on your own audience if it matters. If a color decision is genuinely high-stakes, run your own A/B test with adequate sample size. Somebody else's test on a different page with a different audience tells you very little.
Prioritise accessibility over association. A button that is theoretically the perfect psychological color and fails contrast is worse in every measurable way than a readable one. This is the part of color choice with real, documented, reproducible effects — see our accessible color guide.
The honest summary
Color affects how designs are perceived. It does so mainly through contrast, learned convention and cultural context — not through fixed meanings attached to wavelengths. The effects that replicate are the ones about attention and legibility, which are also the ones you can measure and control.
When you read a confident claim that a specific color produces a specific emotion or a specific conversion lift, ask what population, what context, and whether anyone has replicated it. Usually the answer to the third question is the interesting one.
For the parts of color that are genuinely measurable, our color harmony guide covers building a palette that works structurally, and the contrast checker handles the part with actual numbers behind it.