Human infection with Trypanosoma rhodesiense in the Lambwe Valley area of South Nyanza, Kenya, was first reported in 1959 although T. gambiense sleeping sickness had been the present there since the turn of the century. The two forms of the disease are associated with the presence of the tsetse Glossina fuscipes, which is plentiful and widespread throughout the hinterland, including part of the study area, and G. fuscipes in the thickets near the shores of Lake Victoria and along some of the river systems. Fairly successful attempts have been made to eliminate tsetse by control operations around the lake shore and by aerial spraying of insecticide in the Lambwe Valley. The incidence and geographical distribution of cases of T. rhodesiense sleeping sickness in the Lambwe Valley are described in this report. Eight survey areas were studied between June 1968 and August 1970. Three types of survey procedure were adopted and blood samples were taken for examination. The most reliable method of demonstrating trypanosomes was by direct microscopy of stained blood films. Cases were treated with either suramin or melarsoprol, the latter being preferred for long-standing infections, but a course of suramin was given first. Although there were localized foci of infection in the study area, cases occurred sporadically, usually in April, i.e., in the early part of the long rainy season. It is suggested that, for diagnosis, survey teams should be replaced by a few fixed diagnostic microscopy centres.