Tadpoles have diverse morphologies and occupy diverse habitats. The morphological differences between tadpoles can be represented by linear and geometric measurements and used to explain the organization of tadpole assemblages. However, the effects of evolutionary history must be isolated from the morphological differences before we can determine which patterns result from the use and sharing of common ecological resources. Here, we aimed to determine how morphological similarities and phylogenetic distances affect microhabitat choice by tadpoles. We analyzed the tadpoles of 101 anuran species and classified them according to ecomorphological guild, habitat use, position in the water column, and floor substrate. We used geometric and traditional morphometric approaches to describe the morphological variation among tadpoles and calculated the patristic distance for each species. Afterwards, we used morphometric and phylogenetic matrices as predictors of the variance in the ecological matrix, using a partial redundancy analysis. When we used traditional morphometric data, phylogeny explained a large amount of the ecological variation. By contrast, when we used geometric morphometric data, morphology and phylogeny explained similar amounts of the ecological variation, showing that the technique used to extract morphological variation affects the results. We provide evidence that both morphology, as a surrogate for contemporary factors, and evolutionary inertia are important in determining the behavior of tadpoles. Thus, niche conservatism can be important in modeling the behavior of tadpoles, but does not explain all the preferences of tadpoles.