Is Color Analysis Real? What the Research Says, and What It Doesn’t
Short answer: parts of color analysis rest on things science can measure and has studied, and other parts are conventions that have not been tested. Skin color can be measured precisely, and researchers have shown that the colors around a face change how people judge that face. What has not been done, as far as we could find, is a published study testing whether a seasonal system, or a trained analyst, sorts people consistently and predicts which colors others will prefer on them.
That gap is not the same as evidence against color analysis. It means the question is open. This page sets out what is known, what is not, and how to use color analysis sensibly in the meantime. We wrote it to be fair both to the research and to the many professional analysts who do careful, structured work.
What can be measured: skin color in numbers
Color science has a standard way to put numbers on any surface color, including skin. The CIE (the International Commission on Illumination) defines the CIELAB color space in the standard ISO/CIE 11664-4. It describes a color with three values:
| Value | What it describes | What it means for skin |
|---|---|---|
| L* | Lightness, from 0 (black) to 100 (white) | How light or deep skin is, which is the “depth” or “value” that analysts talk about |
| a* | Green (negative) to red (positive) | Redness, which rises with blood flow, flushing and some skin conditions |
| b* | Blue (negative) to yellow (positive) | Yellowness, which is influenced by pigments such as melanin and carotenoids |
| Hue angle and chroma | Calculated from a* and b*: which direction the color leans, and how strong it is | Closest to the ideas of “warm versus cool” and “bright versus soft,” though not the same thing |
Dermatologists have used these numbers for decades. In 1992, Weatherall and Coombs, in the Journal of Investigative Dermatology, measured forearm skin in healthy volunteers and proposed L*, hue angle and chroma as an unambiguous way to communicate skin color. In 1991, Chardon and colleagues, in the International Journal of Cosmetic Science, used CIELAB readings to propose a skin color classification and to track how tanning moves skin through color space. More recently, Thong, Joniak and Xiang (ICCV 2023) argued that describing skin by lightness alone misses a second dimension, a hue angle running from red to yellow, and recommended measuring both.
So the raw material of color analysis, the lightness and hue of a person’s skin, is measurable and well studied. None of these papers studied undertone labels or seasons, though. Measuring skin is not the same as validating a styling system built on it.
What the published studies tested
A few studies come close to the questions color analysis asks. Here is what each one did, and what it did not do.
| Study | What it tested | What it found | What it did not test |
|---|---|---|---|
| Perrett and Sprengelmeyer, i-Perception, 2021 | Online participants adjusted the color of simulated clothing to suit face photos of 12 White women, six fair-skinned and six tanned | Participants chose cooler hues, such as blues, more often for fair skin and orange-reds more often for tanned skin. Choices tracked skin yellowness (b*) most consistently | Seasonal categories, professional analysts, men, older faces, or a range of ethnic backgrounds. Clothing was shown as flat color shapes |
| Stephen and colleagues, PLoS ONE, 2009 | Participants adjusted the skin color of calibrated face photos to look as healthy as possible | People increased blood-related redness to make faces look healthier, especially the color of oxygenated blood | Clothing colors or any styling system |
| Stephen, Coetzee and Perrett, Evolution and Human Behavior, 2011 | Perceived health from skin yellowness and lightness, and its link to diet | Higher skin yellowness and lightness looked healthier in the faces tested, and carotenoid-like coloring was preferred over melanin-like coloring | Clothing colors or any styling system |
| Lafer-Sousa, Hermann and Conway, Current Biology, 2015 | How 1,401 people described the colors of one photograph of a dress | People split into groups that saw different colors, linked to their assumptions about the light. Adding lighting cues flipped many viewers | Faces or clothing advice |
Read together, these studies support two ideas that analysts rely on. First, small shifts in skin color, toward redder, yellower, lighter or grayer, change how healthy and vivid a face looks to other people. Second, the colors and light around something change how its color is seen. A drape near the face plausibly shifts how the skin appears, and that shift is what an analyst looks for.
The 2021 study is the closest to the actual practice. Its authors noted that the stylists’ personal categories “are not defined scientifically, and advised colours are inconsistent.” They did not test seasonal analysis, though. Their result, that people’s color choices for a face tracked its skin yellowness, is consistent with the warm versus cool idea, but it does not show that sorting people into 4, 12 or 16 groups works.
Simultaneous contrast: why a drape can change your face
Simultaneous contrast is a long-known effect in color perception: a color looks different depending on the colors next to it. A gray square looks lighter on black and darker on white. A neutral patch beside red can look faintly green. The French chemist Michel Eugène Chevreul described these effects in detail in 1839, and textile and paint workers have used them ever since.
Applied to a face, a drape can make skin look slightly redder, grayer or more yellow by contrast, and can make shadows look deeper or softer. This is the most direct link between perception science and what happens in a draping session. It explains why a drape can change your face. It does not by itself say which change most people will judge as better.
Why lighting, screens and phone cameras mislead
- Light source. Warm bulbs add yellow, shade and overcast daylight add blue, and fluorescent light can add green. The same skin and the same fabric will look different under each. The 2015 dress study showed how strongly people’s assumptions about light shape what color they see.
- Automatic white balance. Phone cameras adjust color for each shot to make the scene look “neutral.” Two selfies taken minutes apart can show different skin color, and a bright drape in frame can shift the camera’s correction.
- Filters and beauty modes. These can smooth and warm skin, which changes exactly what an analysis tries to judge.
- Screens. Monitors and phones differ in color and brightness, and night modes warm everything. Two people viewing the same photo may not see the same colors.
This is why careful practitioners insist on controlled light, and why we treat any app that reads your season from a selfie as entertainment rather than measurement.
Which parts are perception and which are convention
| Part of the practice | Grounded in perception or measurement? | Notes |
|---|---|---|
| Skin has measurable lightness and hue | Yes | Well established in dermatology and color science |
| Colors near the face change how skin looks | Yes, in principle | Follows from simultaneous contrast; not tested specifically for drapes |
| Warm versus cool as the first sort | Partly | Consistent with the 2021 finding on skin yellowness and color choices, in a small, narrow sample |
| Four, 12 or 16 groups | Convention | Different systems draw boundaries differently; none has published validation that we found |
| Season names and palettes | Convention | Useful shorthand; palettes differ between companies |
| “Never wear” lists | Convention | A color away from the face, or in a print, behaves differently than a solid at the chin |
How practitioners reduce subjectivity
A draping session is built to reduce the biases that make casual color judgments unreliable. The methods line up well with good practice in any visual comparison:
- Controlled light. Natural daylight or daylight-balanced lamps, the same for every client. House of Colour, for example, performs its analyses in person in natural daylight.
- A neutral frame. Bare skin and a neutral cape, so nothing else is changing.
- Side-by-side comparisons. Two drapes that differ in one quality at a time, such as warmth, while depth stays similar. This isolates the variable being tested.
- A fixed, calibrated drape set. The same fabrics for every client, so results are comparable.
- Repetition. Going back to earlier comparisons to see whether the call holds.
- More than one viewer. Some analysts invite the client, or a second person, to judge, which makes one observer’s preference less decisive.
What would count as better evidence
The questions are testable. Studies that would move the debate include:
- Agreement between analysts. Several trained analysts drape the same people independently, without seeing each other’s results. How often do they agree, within a system and across systems?
- Consistency over time. The same analyst drapes the same people weeks apart. Do results hold?
- Blind ratings. Photos of people wearing in-palette and out-of-palette colors, taken in controlled light, rated by viewers who do not know the palette. Do in-palette photos score better on looking healthy, rested or attractive?
- Diverse samples. Men and women, a range of ages and skin depths, and different backgrounds, unlike the narrow samples above.
- Instrument links. Do measured CIELAB values predict the season an analyst assigns?
Professional analysts are well placed to help with studies like these, since they already work with consistent drapes and light.
A practical way to use color analysis
Our view: use color analysis as a structured comparison tool, not a set of rules. It gives you a method, comparing colors one quality at a time against your face in good light, and a vocabulary of warm and cool, light and deep, bright and soft, for describing what you see.
Nothing on this checklist is saved or sent anywhere.
To learn what happens in a session and how the systems differ, read our guide to professional color analysis. To try a careful home comparison first, start with what is my undertone? and then our color analysis quiz, which is a starting point, not a substitute for a draping. The 12 color seasons pages show the palettes in hex codes.