SummaryA cluster sample survey was conducted in 1998 in 30 schools to assess the effect of the growth of Niamey during the last decade on a urinary schistosomiasis urban focus described in 1989. Two thousand and fortytwo children (11.0 ϩ 0.1 years old) had a urine filtration test and answered a behavioural questionnaire. Snail populations of the sites used by schoolchildren were followed up in 1999. The global prevalence was 15.7% in 1998, as opposed to 23.7% in 1989. The prevalence was very low in schools far from the river and higher in those along the Niger banks, particularly in villages on the periphery of the urban area. Geographical factors were more important than socio-economic ones in explaining the distribution of the disease. Only 46% of the children in Niamey reported water contact; mainly in the river, rarely in pools and the canal. The infection risk was low in pools (RR ϭ 1.6), high in the river (RR ϭ 3.5) and very high in the canal (RR ϭ 12.5). Malacological studies confirmed the location of transmission sites obtained through parasitological studies and the questionnaire. Sixty-one per cent of the children travelled outside Niamey to the hyperendemic surrounding areas. However, these movements did not increase their infection level. The results are discussed in relation to water contact behaviour and Schistosoma haematobium transmission features.