Adult mandibles of 317 modern humans and 91 great apes were selected that showed no pathology. Adult mandibles of Pan troglodytes troglodytes, Pongo pygmaeus pygmaeus and Gorilla gorilla gorilla and from 2 modern human populations (Zulu and Europeans from Spitalfields) were reliably sexed. Thirteen measurements were defined and included mandibular height, length and breadth in representative positions. Univariate statistical techniques and multivariate (principal component analysis and discriminant analysis) statistical techniques were used to investigate interspecific variability and sexual dimorphism in human and great ape mandibles, and intraspecific variability among the modern human mandibles. Analysis of interspecific differences revealed some pairs of variables with a tight linear relationship and others where Homo and the great apes pulled apart from one another due to shape differences. Homo and Pan are least sexually dimorphic in the mandible, Pan less so than Homo sapiens, but both the magnitude of sexual dimorphism and the distribution of sexually dimorphic measurements varied both among and between modern humans and great apes. Intraspecific variation among the 10 populations of modern humans was less than that generally reported in studies of crania (74.3% of mandibles were correctly classified into 1 of 10 populations using discriminant functions based on 13 variables as compared with 93% of crania from 17 populations based on 70 variables in one extensive study of crania). A subrecent European population (Poundbury) emerged as more different from a recent European population (Spitalfields) than other more diverse modern populations were from each other, suggesting considerable morphological plasticity in the mandible through time. This study forms a sound basis on which to explore mandibular variation in Neanderthals, early Homo sapiens and other more ancient fossil hominids.