Several social, environmental, and behavioral variables shown to be strongly associated with the level of Schistosoma mansoni infection were evaluated regarding their power to discriminate infected from noninfected children. Emphasis was placed on the identification of infected persons with high egg counts, who are at risk of developing the severe clinical forms of the disease. Data were collected in a survey carried out in a town (Santo Antonio de Jesus) in northeastern Brazil. The sample consisted of all children 12–14 years old living in the town. An extensive questionnaire was used to collect information on each child and on family and household conditions. Stool samples were taken for quantitative examination. The prevalence of S. mansoni infection was 31.0%. The risk factors were first analyzed one-by-one and then aggregated using logistic models. From the models, individual probabilities of being infected were estimated. The results were then used to produce relative operating characteristics curves. The model with highest discriminative power was that which included five risk factors, namely migratory status and four others related to water-contact activities. The ability to discriminate increased with an increase in the level of egg excretion. The conclusion is that the development of short questionnaires for large-scale screening for S. mansoni infection in endemic areas is realistic.