Discussions of the evolution of sexual dimorphism in torso shape and the pectoral region assume that this dimorphism exists independently of body size. We test this assumption in two human populations and further examine what is needed to understand sexual dimorphism in the pectoral region. Modern human males have broad shoulders and narrow hips relative to females, lending males a more triangular torso. The wider female pelvis is commonly attributed to obstetric pressures while the broader male pectoral girdle has been argued to be an adaptation that improves hunting or intrasexual competition. While sexual dimorphism in the pelvic girdle is known to exist after adjusting for body size across human populations, most studies of sexual dimorphism in the pectoral girdle have not adjusted the data to account for sexual size dimorphism or compared different ancestral groups. The aforementioned hypotheses explaining sexual dimorphism in the clavicle and scapula as products of natural selection are predicated on the untested assumption that sex differences do not scale with body size. This study tests this assumption by comparing various measurements of the pectoral girdle, the pelvic girdle, and six pectoral-pelvic indices of black and white South Africans of known sex and height to test whether the sexes and ancestral groups will differ in these values after adjusting for differences in body size. Comparisons of ancestral groups reveal that white South Africans have larger pectoral and pelvic dimensions than black South Africans, but that blacks have larger index values than whites. Regardless of differences in ancestry and body size, males have significantly broader pectoral regions as indicated by comparisons of both individual pectoral measurements and pectoral-pelvic indices. This pattern of sexual dimorphism is reversed in the pelvic region where females have larger skeletal elements. In addition to finding both absolute and relative differences in mean values for the pectoral and pelvic skeleton, females and males and blacks and whites differ in the scaling relationship of these traits, suggesting different allometric trajectories for these bones that may be explained by their distinct evolutionary functions, their adaptations to specific environments, or by changes in lengths due to age. These results suggest that sexual dimorphism in the pectoral region is not a product of scaling and that differences in this region reflect adaptive forces acting in unique ways on each sex, consistent with the assumptions of earlier evolutionary explanations.