The cranial morphology of the African Old World monkeys Mandrillus, Papio, and Theropithecus (i.e., baboons) has been the subject of a number of studies investigating their systematic relationships, patterns of scaling, and growth. In this study, we use landmark-based geometric morphometrics and multivariate analysis to assess the effects of size, sex, taxonomy, and geographic location on cranial shape. Forty-five landmarks were digitized in three dimensions on 452 baboon crania and subjected to generalized Procrustes analysis (GPA), which standardizes geometric size but leaves scaling-based shape differences in the data. The resulting shape coordinates were submitted to regression analysis, principal components analysis (PCA), partial least-squares (PLS) analysis, and various clustering techniques. Scaling (shape differences correlated with size) was the largest single factor explaining cranial shape variation. For instance, most (but not all) of the shape differences between the sexes were explained by size dimorphism. However, central tendencies of shape clearly varied by taxon (both specific and subspecific) even after variations in size and sex were adjusted out. Within Papio, about 60% of the size-and sex-adjusted shape variations were explained by the geographic coordinates of the specimen's provenance, revealing a stepped cline in cranial morphology, with the greatest separation between northern and southern populations. Based on evidence from genetic studies, and the presence of at least two major hybrid/interbreeding zones, we interpret the phylogeographic pattern of cranial variation as indicating that these populations are best ranked as subspecies of a single species, rather than as two or more distinct biological species. This objective approach can be applied to other vertebrate species or species groups to help determine the taxonomic rank of problematic taxa.