A central problem in ecology is to explain how the great numbers of rare tree species found in tropical rainforests are maintained through time. Approximately 70% of these species occur at a density of less than 2 trees per hectare. Long-term studies of species rarity and forest dynamics using a unique set of three contrasting dipterocarp forests from northwest Borneo have been censused since 1965. Hypotheses which explain the maintenance of rare species are being tested using these data combined with sapling and seedling surveys from 1986 to 1988 and detailed forest maps. One forest is subject to periodic catastrophic drought during which canopy trees die while a second forest is heterogeneous in space due to local landslips. Species richness is greatest in forest which is intermediate in soil fertility and levels of disturbance. Because of these differences among forests, there may not be one simple explanation for the great species richness of these forests. These data are being analyzed jointly with members of the Malaysian Forest Department concerned with practical matters of forest management. The results are also relevant to conservation biologists concerned with the maintenance of biological diversity.