It has often been reported that, in the absence of information about the direction of illumination, people interpret surface convexities and concavities in accordance with the assumption that illumination comes from above. However, 'above' could mean with reference to gravity, the head or the retina. Yonas et al reported that four-year-old infants use the head more than gravity as the frame of reference in interpreting surface relief but that seven-year-olds make about equal use of the two frames of reference. The potency of these two frames of reference when acting separately and when pitted against each other was measured on adult subjects. For all subjects the 'assumption' about the direction of illumination was predominantly with respect to the head. The gravitational frame was used only when the headcentric frame was irrelevant, and then not consistently.