Adaptive hypotheses about colour variation are widespread in behavioural ecology, and several methods of objective colour assessment have been proposed and validated for use in a wide variety of taxa. However, to date, the most objective and reliable methods of assessing colour are not readily applied to wild animals. In the present study, we present a simple method for assessing colour in unrestrained, wild subjects using digital photography. The method we describe uses a digital camera, a colour standard, and colour analysis software, and can be used to measure any part of the visible colour spectrum. We demonstrate that the method: (1) is accurate and precise across different light conditions; (2) satisfies previous criteria regarding linearity and red, green, and blue equality; and (3) can be independently validated visually. In contrast with previous digital methods, this method can be used under natural light conditions and can be readily applied to subjects in their natural habitat. To illustrate this, we use the method to measure chest colour in wild geladas (Theropithecus gelada). Unique among primates, geladas have a red patch of skin on their chest and neck, which, for males, is thought to be a sexually selected signal. Offering some support to this hypothesis, we found differences in chest ‘redness’ for males across different age groups, with males in their reproductive prime exhibiting the reddest chests. © 2008 The Linnean Society of London, Biological Journal of the Linnean Society, 2008, 94, 231–240.