The ecological hypothesis of speciation is that reproductive isolation evolves ultimately as a consequence of divergent natural selection on traits between environments. Ecological speciation is general and might occur in allopatry or sympatry, involve many agents of natural selection, and result from a combination of adaptive processes. The main difficulty of the ecological hypothesis has been the scarcity of examples from nature, but several potential cases have recently emerged. I review the mechanisms that give rise to new species by divergent selection, compare ecological speciation with its alternatives, summarize recent tests in nature, and highlight areas requiring research.