XKCD Survey
In 2010 more than five million colour names were collected from 222,500 sessions of an online survey run by a cartoonist.1
Randall Munroe, who draws the webcomic xkcd, showed visitors colours on screen and asked them to type a name for each.1
He published the results on 3 May 2010.1 Before analysing anything he had to clean the data.1 He excluded people who gave many answers almost nobody else used, and discarded everything from anyone who gave the same name to colours of very different hue.1 Some submissions came from scripts, the most persistent of which entered the same offensive word for 2,400 colours, and a few dozen were attempts to break into the database.1 What remained produced a table of 954 named colours, each tied to an RGB value.1
The survey also asked for each participant's chromosomal sex, because colour blindness is linked to it.1 That made it possible to compare how men and women named colours.1 The headline answer was that they were more alike than different.1 On average both groups named colours similarly, and women used modifying words slightly more often.1
The more striking result came from asking which names each group used out of proportion to the other.1, 2 The names most favoured by women were fashion-style shade names such as dusty teal, blush pink and dusty lavender.1, 2 Those most favoured by men included jokes, insults and a misspelling of beige.1
The linguist Mark Liberman of the University of Pennsylvania wrote two days later that results this neat deserved suspicion, and urged readers to test them against the raw data, which Munroe had released.1, 2 He also pointed out what the survey could not do.2 It could not separate seeing a colour differently from knowing different words for it, or from choosing to use them.2 For some greens and blues, he noted, the differences between the sexes were small.2
A colour has as many names as people willing to type one. The survey showed how often they agree.
References
- Munroe, R. (2010, 3 May). Color survey results. xkcd blog. Link Primary report (survey author's own results) Free to read. Not peer reviewed; it is the survey's own report and the only primary source for its data.
- Liberman, M. (2010, 5 May). Sex and color language. Language Log, University of Pennsylvania. Link Expert commentary (named linguist) Free to read.
Sources last checked 18 October 2026.
Know someone this colour belongs to? Name a colour after them.
Name a colour