Because teeth are commonly preserved in the fossil record, dental remains have often been employed in estimating evolutionary relationships among fossil hominoids. This is appropriate, however, only to the extent that dental morphology is phylogenetically informative. I have used phenetic analytic techniques to assess whether hominoid molars are likely to be useful for phylogenetic inference. Thirty-four occlusal landmarks for first and second molars were chosen; seven on each maxillary and ten on each mandibular tooth. Three-dimensional locations of these points were determined from stereophotographs of dental arcades of more than 260 specimens from six taxa (gorilla, chimpanzee, human, orangutan, siamang, and gibbon). Analytic emphasis was on canonical variates analyses of landmark coordinates for mandibular and maxillary second molars, adjusted for intergroup size differences. There is little correspondence between the systematic implications of hominoid molar morphometrics and reliable estimates of evolutionary propinquity based on interhominoid biomolecular similarities. The former seem to have been determined largely by dietary constraints. Although this suggests the possibility of using the protocol employed here to infer diets of fossil hominoids, molar crown measurements seem unlikely to serve well as phylogenetic indicators in the Hominoidea.