Birth weight as an index of neonatal health needs to be supplemented by other objective measurements, particularly in the case of light-for-dates neonates, where inaccuracies in estimates of gestation confound the problem. Skinfold thicknesses were taken using a Harpenden caliper to obtain further indices of neonatal well-being. The reproducibility of measurements was good with most of the error due to manipulation of the caliper by the operator and not to site selection. In a series of ‘normal’ neonates, sex differences were found, particularly in the flank and quadriceps measurements. These two sites also distinguish best between a small group of light-for-dates subjects and the ‘normal’ group.