Synopsis Several methods exist for measuring cutaneous relief. Profilometry allows the measurement of the profile of a skin-positive araldite replica. Image analysis detects and measures shadows behind the crests created by glazing lighting on a negative silicone skin replica. We have studied the correlation between both methods for twenty-nine crow's feet replicas. The area explored by image analysis is a circle of fifteen millimeters diameter. The areas explored by profilometry are, on one hand, ten scans 9500 mum long at intervals of 60 mum, i.e., 5.13 mm(2) (D 60 mum), and on the other hand ten scans of the same length at intervals of 900 mum, i.e., 76.95 mm(2) (D 900 mum). All parameters of peak height and surface were correlated. The best correlations were obtained between image analysis and profilometry for D 900 mum. In this case the P parameter (wrinkles depth measured by image analysis) was influenced for 46% by surface S under the peaks and for 33.5% by the peak height H measured by profilometry. In the same way the coefficient of developed skin surface CDSS was influenced for 45% by H and for 44% by S. The field of analysis common to both methods was found to be middle wrinkles with depth between 45 and 100 mum.