Consult your doctor. These are research instruments, not tests of health. They measure what a person reports or does on one occasion, against a stated reference group. Nothing here diagnoses, treats, cures or prevents any disease, and a score is not a finding about your health. Take the record to a clinician if you want it interpreted — that reading is theirs to make, not ours.

Coloured light is presented here as history and as published research, not as a treatment. This is not a medical device and not medical advice. Nothing here diagnoses, treats, cures or prevents any disease, no colour is matched to a condition, and no claim is made that light measures or restores anything about your body. Where the research is strong we say so; where the tradition’s claims are unsupported we say that too.

This is not a laboratory result. No specimen was taken and no clinical assay was run. It is a record of self-administered measurements — the same kind of thing as a sleep diary or a home blood-pressure log — laid out so a clinician can read it quickly.

What you call a colour

Your own colour lexicon: we show a colour, you type what you call it.

There are no right answers here

We show you a colour. You type what you call it. One word or several, whatever you would actually say — “blue”, “teal”, “that green my aunt’s kitchen was”. We keep exactly what you type. Nothing is corrected, nothing is snapped to a list of approved colour words, and the unusual answers are the point rather than the noise.

40 colours, about four minutes. You can keep going for as long as you like — people do — and you can stop whenever you want.

This result will never be worded as: “This tells you something about your eyes.” — naming is a cultural and linguistic trait, not a retinal one, and this page is not a vision test.

Your key

Your browser made you a random key; we store a one-way hash of it and never the key. Write it down if you want to come back or delete this later.

…

Already have one?

Which language are you naming in?

This changes what the answers mean and nothing else. It is not required.

Language you are naming in Other languages you speak

The colours

What this page can and cannot do

It can do this one honestly.The question is about the mapping from appearance to word, and both sides of that mapping are sampled inside one observer on one screen. Whatever your display does to a colour, it does to the colour you are naming — so the errors that wreck most online colour tests largely cancel here. This exam is native to the medium: the xkcd colour survey collected roughly 3.4 million responses this way on uncalibrated monitors.
It cannot guarantee two people saw the same colour.They did not, quite. Pooling answers across people needs the display’s gamut class as a covariate, and comparing them directly needs matched hardware.
It is not a vision test and cannot become one.A colour-vision deficiency is diagnosed with plates and an anomaloscope by a clinician, not by a text box.

Where this comes from

Lindsey & Brown (2014), Journal of Vision 14(2):17 — the colour lexicon of American English.
Berlin & Kay (1969), Basic Color Terms.
Munroe (2010), the xkcd Color Survey — crowdsourced, blog-published, not peer-reviewed.
Zaslavsky et al. (2019), Topics in Cognitive Science — naming reflects communicative need, not only perceptual structure.

The rules this page is bound by

Report the measurement, not the category."Your VVIQ score was 19 out of 80. Scores at or below 24 are the range in which most published aphantasia research recruits participants." Never "you have aphantasia."
Name the reference class, every time.A percentile is meaningless without saying among whom. Ours is: among people who found this page and chose to take the test. That group is systematically enriched for exactly the traits being measured, because people who suspect something about themselves go looking.
Consistency, not confidence.For every instrument the second administration is part of the instrument. A one-shot result is a reading, not a finding.
Never route a result to anything consequential.No credential tier, no karma weight, no grant eligibility, no payment. The moment a score buys something, people optimise it and the data dies.