Explaining variation in species richness among provinces and other large geographic regions remains one of the most challenging problems at the intersection of ecology and evolution. Here we argue that empirical evidence supports a model whereby ecological factors associated with resource availability regulate species richness at continental scales. Any large-scale predictive model for biological diversity must explain three robust patterns in the natural world. First, species richness for evolutionary biotas is highly correlated with resource-associated surrogate variables, including area, temperature, and productivity. Second, species richness across epochal timescales is largely stationary in time. Third, the dynamics of diversity exhibit clear and predictable responses to mass extinctions, key innovations, and other perturbations. Collectively, these patterns are readily explained by a model in which species richness is regulated by diversity-dependent feedback mechanisms. We argue that many purported tests of the ecological limits hypothesis, including branching patterns in molecular phylogenies, are inherently weak and distract from these three core patterns. We have much to learn about the complex hierarchy of processes by which local ecological interactions lead to diversity dependence at the continental scale, but the empirical evidence overwhelmingly suggests that they do.