Several recent studies have stressed the role of dietary change in the origin and early evolution of our genus in Africa. Resulting models have been based on nutrition research and analogy to living peoples and nonhuman primates or on archeological and paleoenvironmental evidence. Here we evaluate these models in the context of the hominin fossil record. Inference of diet from fossils is hampered by small samples, unclear form-function relationships, taphonomic, factors, and interactions between cultural and natural selection. Nevertheless, craniodental remains of Homo babilis, H. rudolfensis, and H. erectus offer some clues. For example, there appears to be no simple transition from an australopith to a Homo grade of dietary adaptation, or from closed forest plant diets to reliance on more open-country plants or animals. Early Homo species more likely had adaptations for flexible, versatile subsistence strategies that would have served them well in the variable paleoenvironments of the African Plio-Pleistone.