A study to determine the diagnostic performance of simple anamnestic questions and recalled water-contact patterns for self-diagnosis of Schistosoma mansoni infection was carried out in western C te d'Ivoire. A total of 322 schoolchildren were screened over four consecutive days with the Kato-Katz technique to assess S. mansoni and concurrent geohelminth infections. Children were individually interviewed by teachers using a standardized questionnaire asking about symptoms, reported diseases, and water-contact patterns. The cumulative infection prevalence of S. mansoni was 76.4%. Univariate statistics revealed a significant association between the level of S. mansoni infection and three recalled water contact patterns: (1) fishing with nets, (2) swimming/bathing and (3) crossing rivers, but no significant association with reported symptoms and/or reported diseases. Multivariate analysis revealed significant adjusted odds ratios (OR) for crossing the river Tchéorbour (OR = 3.90, P = 0.007), crossing the river Sonbour (OR = 3.90, P = 0.008) and swimming/bathing in the latter (OR = 3.28, P = 0.017). The diagnostic performance of these water-contact patterns was characterized by high specificities but low sensitivities, hence negative predictive values. In the village studied here, recalled water-contact patterns were more useful variables than anamnestic questions for schoolchildren's self-diagnosis of S. mansoni infection, but no generalization of these findings beyond this population is possible at this time.