We tested the ability of sound recordings relative to that of point counts to estimate species richness in the Tambopata Reserve in southeast Peru. We tested the effect of two environmental factors (estimated richness and presence of noisy species) and two attributes of species (abundance and foraging height) on estimates of species richness made by point counts and sound recordings. Sound recordings are preferred to point counts when richness is high, as during the dawn chorus, because they allow for repeated listenings. Point counts are more effective than sound recordings at detecting rarely heard species. The presence of noisy species at a station had no effect on the relative ability of the two methods to measure species richness. The foraging height of a species had no effect on its relative detectability by either method. Sound recording was found to be a suitable alternative to point counts for estimating species richness and a preferable alternative under some circumstances.