The Lower Kihansi Hydropower Project in southern Tanzania caused the diversion of the Kihansi River from the Kihansi Gorge in the year 2000. By sampling the understorey avifauna prior to diversion, we examined (i) whether the adjacent Udagaje Gorge was an adequate control for observations in the Kihansi Gorge; (ii) which species of conservation interest occurred; and, (iii) which season best suited annual monitoring. Species composition and capture rates at three and two elevational transects in the Kihansi Gorge and Udagaje Gorge, respectively, confirmed that Udagaje had a comparable avifaunal assemblage to Kihansi. The cold season was most appropriate for population monitoring because >2 times more individuals were captured in the cold than hot season at both gorges, and at least four altitudinal migrants were present in the cold but not hot season. Post-diversion sampling revealed that only the Upper Kihansi transect suffered a significant decrease in number of individuals, a result that was driven largely by a decline in the Little Greenbul, Andropadus virens. This transect is closest to the Kihansi waterfall and associated spray zone which were lost after river diversion. Lack of differences in bird communities at other transects after diversion illustrates that early post-diversion effects on birds are probably concentrated near the base of the main falls. Together with studies of other biota in Kihansi, we propose that longterm monitoring is necessary to understand the factors that regulate changes in species composition of this threatened forest site.