In numerous studies on colour constancy or colour induction subjects have to adjust a test field so that it looks achromatic. Their viewing behaviour during these settings is often not controlled or reported. Here I show that the results of grey settings depend on whether subjects visually explore the stimulus by looking around or fixate their gaze exclusively on the test field. Two different viewing instructions are compared with regard to the degree of constancy as measured by the shift of grey settings in coloured surrounds. In variegated surrounds (but not in uniform surrounds) there is a robust effect of viewing condition for all subjects and all surround chromaticities tested, in that exploration increases colour constancy compared to fixating the test field. Values of a colour constancy index are increased by as much as 20% (where 100% colour constancy means that the subject chooses as grey the mean chromaticity of the surround) with an average across all subjects and surrounds of 12.6%. Thus, if this factor is not experimentally controlled, it could inflate variance, reduce comparability between different studies, and even lead to unwarranted conclusions if viewing behaviour unpredictably differs between experimental conditions.