<p>Colour constancy is standardly defined as the visual phenomenon in which an object’s colour tends to be perceived as (approximately) constant regardless of (drastic) variations in the illumination and other viewing conditions. But a closer look at our colour experience reveals that the colour initially held to be constant varies successively and simultaneously. If colour constancy is a genuine form of perceptual constancy, then there seems to be an ambiguity involved in it that needs to be explained. In this article, I explore Merleau-Ponty’s contributions to this ambiguity problem from his 1945 work, the <i>Phenomenology of Perception</i>. In the first part, I argue that Merleau-Ponty’s descriptions of our colour experience expose the inadequacies of two lines of response to the ambiguity problem proposed by the computational approach to colour constancy, namely the “imperfection response” and the “other-aspect response”. In the second part, I reconstruct Merleau-Ponty’s theory of colour constancy, which contains three main theses: (i) the pluralist thesis, (ii) the normative thesis, and (iii) the holistic thesis. I suggest that each of these theses provides an explanation that can do justice to the ambiguity of colour constancy.</p>