Ants are the world's most diverse and ecologically dominant eusocial organisms. Resolving the phylogeny and timescale for major ant lineages is vital to understanding how they achieved this success. Morphological, molecular, and paleontological studies, however, have presented conflicting views on early ant evolution. To address these issues, we generated the largest ant molecular phylogenetic data set published to date, containing approximately 6 kb of DNA sequence from 162 species representing all 20 ant subfamilies and 10 aculeate outgroup families. When these data were analyzed with and without outgroups, which are all distantly related to ants and hence long-branched, we obtained conflicting ingroup topologies for some early ant lineages. This result casts strong doubt on the existence of a poneroid clade as currently defined. We compare alternate attachments of the outgroups to the ingroup tree by using likelihood tests, and find that several alternative rootings cannot be rejected by the data. These alternatives imply fundamentally different scenarios for the early evolution of ant morphology and behavior. Our data strongly support several notable relationships within the more derived formicoid ants, including placement of the enigmatic subfamily Aenictogitoninae as sister to Dorylus army ants. We use the molecular data to estimate divergence times, employing a strategy distinct from previous work by incorporating the extensive fossil record of other aculeate Hymenoptera as well as that of ants. Our age estimates for the most recent common ancestor of extant ants range from approximately 115 to 135 million years ago, indicating that a Jurassic origin is highly unlikely.