Sharks are a diverse and ecologically important group, including some of the ocean's largest predatory animals. Sharks are also commercially important, with many species suffering overexploitation and facing extinction. However, despite a long evolutionary history, commercial, and conservation importance, phylogenetic relationships within the sharks are poorly understood. To date, most studies have either focused on smaller clades within sharks, or sampled taxa sparsely across the group. A more detailed species-level phylogeny will offer further insights into shark taxonomy, provide a tool for comparative analyses, as well as facilitating phylogenetic estimates of conservation priorities. We used four mitochondrial and one nuclear gene to investigate the phylogenetic relationships of 229 species (all eight Orders and 31 families) of sharks, more than quadrupling the number of taxon sampled in any prior study. The resulting Bayesian phylogenetic hypothesis agrees with prior studies on the major relationships of the sharks phylogeny; however, on those relationships that have proven more controversial, it differs in several aspects from the most recent molecular studies. The phylogeny supports the division of sharks into two major groups, the Galeomorphii and Squalimorphii, rejecting the hypnosqualean hypothesis that places batoids within sharks. Within the squalimorphs the orders Hexanchiformes, Squatiniformes, Squaliformes, and Pristiophoriformes are broadly monophyletic, with minor exceptions apparently due to missing data. Similarly, within Galeomorphs, the orders Heterodontiformes, Lamniformes, Carcharhiniformes, and Orectolobiformes are broadly monophyletic, with a couple of species 'misplaced'. In contrast, many of the currently recognized shark families are not monophyletic according to our results. Our phylogeny offers some of the first clarification of the relationships among families of the order Squaliformes, a group that has thus far received relatively little phylogenetic attention. Our results suggest that the genus Echinorhinus is not a squaliform, but rather related to the saw sharks, a hypothesis that might be supported by both groups sharing 'spiny' snouts. In sum, our results offer the most detailed species-level phylogeny of sharks to date and a tool for comparative analyses. © 2010 Elsevier Inc.