Morbidity due to infection with Schistosoma mansoni was investigated in a recently discovered highland focus around Lake Cohoha, Burundi. The distribution of the infection was very focal and morbidity patterns in populations from an endemic area A (prevalence 38%, mean egg load of positive subjects 231 eggs per gram [epg]), a less affected area B (16%, 90 epg) and a virtually non-endemic area C (5%, 45 epg), were compared; apart from schistosomiasis, the profiles of these populations were highly similar. The overall frequencies of diarrhoea were 36%, 25%, and 19%, respectively; of abdominal pain 86%, 78%, and 83%; of fatigue 7%, 2%, and 1%; of left lobe hepatomegaly 30%, 18%, and 9%; of right lobe hepatomegaly 18%, 10%, and 5%; of splenomegaly 18%, 10%, and 7%. Organomegaly was generally mild, even in area A. Within area A, the association between the presence of infection and diarrhoea, fatigue, hepatomegaly and splenomegaly was significant, but far less impressive than the results of the community-based comparison with areas B and C. The correlation with intensity was limited to an increased prevalence of diarrhoea and fatigue in the highest egg count group, and a more gradual increase (varying with age) in hepatomegaly and splenomegaly. The data are compared to other morbidity studies in subsaharan Africa, in particular one in the nearby Rusizi Plain. The lesser impact of malaria, the higher egg loads, the recent establishment of the focus and possibly parasite strain differences may account for the more apparent and more important schistosomiasis morbidity in the Cohoha focus. It is concluded that morbidity due to S. mansoni infection is better demonstrated by comparing endemic with non-endemic communities than by relating morbidity to egg counts. Individual egg counts are not always a reliable measure for actual pathology, and their use as a discriminative screening tool for treatment may therefore be questioned.