This study sets out to document and analyse sexual dimorphism in the teeth and bones of five primate groups, including man. Specimens were only included in the analysis if their sexual attribution was reliable and was based on non-osteological criteria. Ninety raw measurements, both cranial and post-cranial, were used and 11 indices were computed from them.
The parameters of each sample were computed in order to compare these results with previous estimates of dimorphism which have not always been based on reliably sexed samples. Correlation matrices were computed.
The overall sex differences were subdivided into “shape” and “size” components using Penrose's size and shape distances and by computing the principal components of each data set. The visually apparent shape differences were confirmed metrically and then examined to see whether they could be explained by allometric effects or whether there was evidence for sexual differences in growth patterns.
Using femur length as the independent variable, because of its correlation with overall size, allometric coefficients were computed for the logarithmically transformed data. The coefficients were in some cases very different between sexes but the majority did not achieve statistical significance. Of those that were significantly different only in Homo did such differences narrowly exceed the number that would be expected by chance alone.
As the vast majority of allometric coefficients for the pooled male and female data differed from unity, the hypothesis that most of the considerable shape differences that exist between some male and female primates are due to underlying growth differences must be rejected. It is suggested that such differences are simply the result of disproportionate change in size.