The prevalence of urinary schistosomiasis among schoolchildren in Pangani District (Tanzania) was assessed rapidly by a questionnaire approach. Based on the results, a strategy of selective treatment with praziquantel was adopted. Eleven primary schools in Mwera Division, Pangani District, with about 2500 schoolchildren were included in a control programme for urinary schistosomiasis. Macro- and microscopic haematuria diagnosed visually and with urine reagent strips was used as an indirect indicator of Schistosoma haematobium infection. Intensity of infection among children was monitored in class 5 (median age 14 years, range 11-17) by urine filtration techniques. Treatment was administered as 40 mg/kg praziquantel in a single dose at the beginning of the school year. The programme was implemented by schoolteachers and coordinated by the District Health Management Team in collaboration with the District Education Office. Teachers were responsible for carrying out all programme activities. Community participation was through collaboration with Teachers-Parents Associations and Village Health Committees. Coverage at yearly (1995-99) examination varied from 67.7% to 80.3%. Prevalence of haematuria decreased from 51.2% (range 22.2-89.5%) at baseline to 23.4% (range 5.8-56.7%) in 1999, a reduction of 54.3%. Macrohaematuria was 21.2% at baseline and 7.2% in 1999, a reduction of 66.0%. Prevalence of infection in class 5 was reduced by 71.4% and geometric mean intensity of positives reduced from 71 eggs/10 mL (95% confidence interval [CI] 52.5-97.7) to 28 eggs/10 mL (95% CI 25.7-55.0), a reduction of 60.6%. Teachers were highly committed, and secured community participation and a smooth implementation of the programme. The community accepted the introduction of a cost-recovery system, whereby parents pay for the treatment of children with episodes of visible haematuria during the school year. Communities also participated in the improvement of sanitary installations at the schools.