The molecular evolution of protein coding sequences has been studied for nearly half a century. The product of this research has been the development of probabilistic models to make evolutionary inferences based on extant sequences and a theoretical framework within which to test these models, including the neutral theory and the molecular clock hypothesis. In comparison, we know relatively little about the evolution of cis-regulatory sequences, whether the same models apply, and whether similar conclusions can be reached. Towards filling this void, we are generating probabilistic models for the evolution of transcription factor binding sites, applying these models to polymorphism and divergence data and testing the predictions of the models using gene expression assays. The ultimate goal is a general understanding of the evolution of cis-regulatory sequences, including the frequency of changes in gene regulation and the mechanism by which such changes occur.