A procedure is developed and applied to evaluate alternative explanations for morphological patterns in communities of Darwin's ground finches. The first step in the procedure is the computation of expected population density for a hypothetical solitary finch species on an island, as a function of beak depth. This was done for 15 Galapagos islands where food characteristics have been measured. The second step involves construction of hypothetical finch communities for these islands using five different models. Models differ in the extent to which processes of assembly and/or evolution favor species of high expected density, and in the extent to which interspecific competition influences these processes. By comparing predictions of models to actual communities, the roles of food supply and competition could be assessed. Results reveal that expected density is usually a polymodal function of beak depth. Islands differ substantially in the shapes of their density functions. Mean beak sizes of species actually present on each island correspond to local maxima in expected density. However, two species never occupy the same or closely adjacent local maxima. Simple models incorporating the effects of both food supply and interspecific competition on assembly/evolution are shown to accurately predict observed morphological patterns. The results support the hypothesis that both food supply and interspecific competition have determined morphological properties in communities of these finches.