According to models of adaptive radiation and sexual selection trait divergence is a key step in the speciation process. At a macroevolutionary scale this leads to the prediction that speciation is driven by the rate of trait evolution rather than the state of the trait. Despite the established theoretical links between trait evolution and speciation, most macroevolutionary studies focus on predicting speciation as a state dependent process. To date the hypothesis that rapid speciation is driven by rapid trait evolution has rarely been tested. If the rate or degree of divergence in species traits is an important factor in driving differences in speciation rates among lineages then there must also be variation in rates of trait evolution among lineages. Using my recently completed phylogeny of all extant birds, novel datasets of ecomorphological and secondary sexual traits, and computer simulations, I will carry out the first large-scale analyses to determine the extent to which rates of phenotypic evolution vary among lineages and across space, test the causes of non-constant evolutionary rates, and derive and test predictions on the effects of rates of trait evolution on speciation.