Ecologic modeling of Schistosoma transmission in endemic communities has suggested that antiparasite therapy targeted at the most heavily infected segment of the human population (i.e., school-age children) should have a significant impact on local parasite transmission. Our 1984-1991 experience with age-targeted therapy in the Msambweni area of Kenya has shown an overall decrease in area transmission within 1-2 years following initiation of annual treatment of school-age groups. Snail monitoring confirmed a continuing but variable reduction of vector infection rates. However, subgroup analysis showed significant differences in transmission suppression between more developed coastal villages with piped-water kiosks and villages with only limited access to safe water supplies. Villages without piped water were marked by higher initial prevalences of S. haematobium infection, greater prevalence among adults, longer and more frequent contact with high-risk water sources, and persistently high transmission despite compliance with parasitologic screening or drug therapy. We conclude that targeted therapy had a significant impact on S. haematobium transmission in some areas, but that more extensive or more prolonged coverage is necessary to reduce the rate of new infection in high-risk villages. Defining field-use algorithms, based on decision analysis of economic and ecologic parameters, should provide effective guidelines for selective versus mass treatment in expanded control areas.