Deciphering relationships among the orders of placental mammals remains an important problem in evolutionary biology and has implications for understanding patterns of morphological character evolution, reconstructing the ancestral placental genome, and evaluating the role of plate tectonics and dispersal in the biogeographic history of this group. Until recently, both molecular and morphological studies provided only a limited and questionable resolution of placental relationships. Studies based on larger and more diverse molecular datasets, and using an array of methodological approaches, are now converging on a stable tree topology with four major groups of placental mammals. The emerging tree has revealed numerous instances of convergent evolution and suggests a role for plate tectonics in the early evolutionary history of placental mammals. The reconstruction of mammalian phylogeny illustrates both the pitfalls and the powers of molecular systematics.