It is commonly assumed that faunal diversity reaches a maximum in tropical rain forest. At this end of the diversity scale, we expect an increase in the number of interactions between ecologically similar species, which should increase the pressure for ecological segregation. Arguing from a different basis, Margalef (1963) suggests that the relationships between species become more specialized, that stenophagy and other precise relationships increase in these rich and complex communities. A consequence of these suggested relations is that niches should tend to be discrete and to overlap only slightly in rain forest communities. In contrast, Klopfer & MacArthur (1961) maintain that the niches of sympatric species of tropical birds overlap more than do niches of temperate zone birds. Their conclusion is based on the ratios of culmen lengths*, which in turn are used as a measure of feeding behaviour. At best this can be considered only a very limited characterization of the niches occupied by the birds in question. Furthermore, the kinds of tropical environments studied by Klopfer and MacArthur were not specified. Consequently, we still do not know the extent to which the niches of related, coexisting species overlap in tropical rain forest. If niches were completely separated, occasional restriction of certain resources or occasional fluctuation in the physical qualities of an environment might result in the elimination of some niches. Overlapping niches, on the other hand, might provide the system with the flexibility needed to resist such perturbations. The degree of overlap of niches of related species is, therefore, ecologically significant. Observations made by our field party in Sarawak, Borneo in 1962 and 1963 provide information bearing on this matter of niche overlap. The hilly tropical rain forest of our work area is laced with numerous small, permanent streams. Many species of frogs and toads live along or in these streams and are easily collected or observed in numbers by persons wading up the streams at night. Frogs have another advantage as subjects for ecological study: they can be caught by hand and their positions recorded with precision. There are three serious limitations in dealing with niches. The first is that niches cannot be completely defined; there will always remain some unknown dimensions. Secondly, we still lack a satisfactory, rigorous way of measuring niche dimensions. Thirdly, what is true of one group of animals may not be true of another; the kinds of specializations open to insects apparently are not available to most vertebrates. In this analysis, characterization of niches depends on eight main attributes: size of animal; position on stream (along course of stream, distance from stream bed and height