A long-term schistosomiasis control programme was initiated on Pemba Island, Tanzania in 1986 with the aim of eliminating morbidity due to Schistosoma haematobium infection. The programme used haematuria as a community indicator of morbidity for surveillance and for identifying individuals for selective population chemotherapy. Analysis of the first 3 evaluation studies showed that the prevalence of microhaematuria was linearly related to the prevalence of infection, and that visually detectable haematuria was a marker of the intensity of infection and risk of morbidity in a community. These relationships remained consistent during repeated community-based chemotherapy, suggesting that measures of haematuria may be useful tools for surveillance.