Early hominins present an unusual pattern of sexual dimorphism. On one hand, the canine teeth of these species are weakly size-dimorphic, vertically short, and nonhoning, suggesting a social system characterized by infrequent, low-intensity intermale competition and monogamous pair-bonding. On the other hand, marked size variation in skeletal remains attributed to species of Australopithecus is thought to reflect strong body-mass dimorphism, which is more consistent with intense intermale competition. Reconciling these conflicting signals and understanding their adaptive significance is a major goal of paleoanthropology. This dissertation research contributes to this objective by investigating factors that may constrain or reduce canine height in extant anthropoid primates. Two hypotheses regarding the relationship between canine height and other elements of the masticatory system were tested using phylogenetic comparative methods. According to the first hypothesis, canine reduction is a pleiotropic by-product of changes in the sizes of other components of the dentition. With respect to canine height, the results of this study fail to support this idea. There is limited evidence for a relationship between basal canine crown dimensions and incisor and postcanine size, but significant interspecific correlations between these variables are not strong and are restricted primarily to the female maxillary dentition. These results indicate that if pleiotropy influences canine size, then its effects are weak. The second hypothesis proposes that canine reduction is a consequence of selection for increased jaw-muscle leverage. This hypothesis receives some support: there is a clear inverse relationship between canine height and the leverage of the masseter muscle in male anthropoids. Females do not exhibit this association due to the fact that dimorphism in muscle leverage is weak or absent in most anthropoid species; in other words, female muscle leverage tracks male muscle leverage, which is linked to canine height. Leverage of the temporalis muscle is not correlated with canine height in either sex. Two specimens of the 3.0-3.7-million-year-old hominin Australopithecus afarensis fall at or beyond the upper end of the great ape range of variation in masseter leverage, which is consistent with the idea that hominin canine evolution was influenced by selection for increased jaw-muscle leverage.