Infants have asymmetrical exposure to different types of faces (e.g., more human than nonhuman, more female than male, and more own-race than other-race). What are the developmental consequences of such experiential asymmetry? Here, we review recent advances in research on the development of cross-race face processing. The evidence suggests that greater exposure to own- than other-race faces in infancy leads to developmentally early differences in visual preferences for, recognition of, formation of categories for, and scanning of own- and other-race faces. Further, such perceptual differences in infancy may be associated with the emergence of implicit racial bias, consistent with a perceptual-social linkage hypothesis. Current and future work derived from this hypothesis may lay an important empirical foundation for the development of intervention programs to combat the early occurrence of implicit racial bias.