A population-based clinical epidemiologic study on schistosomiasis mansoni was carried out in Tuparecê, Minas Gerais. The patients were interviewed for symptoms, water contact, past history and examined for spleen and liver enlargement. From the 830 people registered in the census, 777 (93.6%) had their stools examined (Kato-Katz method) and 696 (83.9%) were clinically evaluated. The overall index of Schistosoma mansoni infection was 43.2%. Significant and increased infection risks could be detected in the young age group (2-14 years old) regarding occupation, time of residence in the area and frequency of water contact. Bloody stools were significantly more prevalent among positives, while diarrhea was significantly more prevalent among those negative. The area was shown to have a low morbidity as well as a low intensity of infection measured by the number of S. mansoni eggs per gram of feces. A close correlation was found between water contact pattern and the age prevalence curve. It has emphasized the importance of habits in determining prevalence rates, besides suggesting that schistosomiasis mansoni in the area is manifested as a light and somewhat harmless infection with little consequence for the population as a whole.