Niche partitioning is a long-standing but heavily debated hypothesis to explain the maintenance of diversity in tropical forests. In particular, suggestions that species are differentially specialized in their germination requirements or in their growth and survival in specific environmental conditions make testable predictions for processes occurring at the early life stages of trees. I examined five years of tree seedling dynamics in Yasuní National Park, Ecuador, to determine whether species demonstrate significant habitat associations measured by significant high or low abundance, growth rate, or survival rates. For these associations to contribute to the maintenance of diversity, species' habitat associations must differ significantly from assemblage-wide patterns to indicate species are partitioning the habitat. For each habitat I tested the observed performance of seedlings of over 40 woody species against a null distribution of seedling performance calculated using 5000 randomized topographic maps produced by torus translations of the topography. Many of the species show significant positive or negative associations with particular habitats in their abundance, growth or survival, and these associations are consistent across years. Several species also show performance differences across habitats that do not match the patterns of the overall seedling assemblage, lending evidence to the importance of habitat partitioning. The majority of species examined displayed neutral associations with habitats, however. While this evidence confirms the existence of habitat associations by seedlings, the frequency of these associations is unlikely to explain the high levels of alpha diversity seen in lowland tropical rainforests.