Sympatric speciation, the evolution of reproductive isolation without geographic barriers, remains highly contentious. As a result of new empirical examples and theory, it is now generally accepted that sympatric speciation has occurred in at least a few instances, and is theoretically plausible. Instead, debate has shifted to whether sympatric speciation is common, and whether models assumptions are generally met in nature. The relative frequency of sympatric speciation will be difficult to resolve, because biogeographic changes have obscured geographical patterns underlying many past speciation events. In contrast, progress is being made on evaluating the empirical validity of key theoretical conditions for sympatric speciation. Disruptive selection and direct selection on mating traits, which should facilitate sympatric speciation, are biologically well supported. Conversely, costs to assortative mating are also widely documented, but inhibit speciation. Evaluating the joint incidence of these key factors may illuminate why sympatric speciation appears to be relatively uncommon.