Short answer
Colour reliably changes which flavour people expect and often which flavour they report, but its effect on how strong or sweet something tastes is inconsistent between studies. The best-supported account is expectation: colour sets a prediction, a small mismatch pulls the experience towards it, and a large mismatch can be noticed and rejected. The associations themselves differ between people and regions.
The laboratory evidence splits in two. When the colour of a drink is changed, people's identification of its flavour shifts with it: a drink coloured to suggest a different fruit is more often identified as that fruit. That result has been replicated many times. Effects on intensity — how sweet, how strong — are a different matter. Charles Spence's 2019 review of the field describes the intensity results as mixed, with significant effects in some studies, none in others, and sometimes both within one study; one experiment found added colour raised the rated intensity of a smell sniffed from the glass while lowering it when the same aroma reached the nose from the mouth. Any claim that a given colour 'makes food taste sweeter' by a fixed amount outruns this evidence.
The explanation with most support is that colour sets up an expectation and the brain combines it with what the tongue and nose report. When the gap is small, the experience moves towards the expectation — assimilation; a dark purple drink expected to be blueberry that is actually blackcurrant tends to be rated in line with the expectation. When the gap is large, the mismatch is noticed and the visual cue is discounted or the product seems worse than it would have without the misleading colour — contrast. Piqueras-Fiszman and Spence's review frames this in predictive-coding terms: cues outside the food, including packaging, labels and price, feed the prediction. Nobody yet knows where the switch from assimilation to contrast happens, which is why colour cannot be tuned to deliver a chosen taste.
For packaging, the practical consequence is modest but real: a pack colour that sets the wrong flavour expectation can make a good product disappoint, because the mismatch is what is tasted.
Packaging colour is one step further removed from the food than the food's own colour, but it feeds the same expectation. The review by Spence and Velasco summarises a growing body of experiments in which the colour of a pack or container changed expected and sometimes perceived taste, and even the perceived fragrance of home and personal-care products. The studies are mostly small, many use drinks or snacks in laboratory conditions, and the effect sizes vary widely. The safe reading is that pack colour is part of the flavour cue a shopper carries to the first taste, not a lever with a known size.
Colour–flavour associations are learned, and they vary. In one study reported in the 2019 review, most young Taiwanese participants expected a clear blue drink to taste of mint, while most young British participants expected raspberry. A large museum-based study with more than five thousand visitors found red was the colour most often chosen as sweetest in every region, but with its share ranging from about 37% of Asian participants to about 51% of South American ones, while blue ranged from about 17% to about 33%. Even within one country, people do not all agree. Colour is a strong cue precisely because it is learned from the products around us, which also makes it a local one.
| Colour | All participants | Lowest regional share | Highest regional share |
|---|---|---|---|
| Red | 40.7% | 37.2% (Asia) | 51.4% (South America) |
| Blue | 27.8% | 17.0% (Asia) | 32.6% (UK) |
Why: The new pack colour signals a different flavour or strength, and the mismatch is experienced as a worse product.
Fix: Test taste with the actual pack in view, not blind, and compare against the old pack.
Why: Colour–flavour matches are local; the colour chosen may point to a different citrus there.
Fix: Check local associations with a short matching task before fixing variant colours.
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.
FactStrong evidence
Changing the colour of a drink reliably shifts which flavour people identify, but the effect of colour on rated flavour intensity has been inconsistent across studies.
Caveat: Mostly laboratory studies with coloured drinks; the review notes these bear little relation to products on real shelves.
Source: On the Relationship(s) Between Color and Taste/Flavor (Spence, Experimental Psychology, 2019)
FactStrong evidence
Expectations set by cues outside the food, such as packaging, labels and price, measurably affect sensory perception, liking and consumption.
FactModerate evidence
In one study most young Taiwanese participants expected a transparent blue drink to taste of mint, whereas most young British participants expected raspberry.
Measured: Young adult participants in Taiwan and the UK.
Caveat: Reported in a review; one study with student-age participants, not a statement about either population as a whole.
Source: On the Relationship(s) Between Color and Taste/Flavor (Spence, Experimental Psychology, 2019)
Colourwise interpretationModerate evidence
Because the point at which a colour–flavour mismatch switches from assimilation to contrast is unknown, a pack colour that sets a clearly wrong flavour expectation is a risk to the product experience rather than a way to tune taste.
Based on: Colourwise's reading of the assimilation/contrast account and the review's statement that the size of discrepancy needed for the transition is not known.
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.