This paper describes a rapid, simple, cost-effective questionnaire for screening school-aged children at risk for Asian schistosomiasis in China. Five hundred and thirty-two children, aged 8–14 years, were selected from 3 schools in an area moderately endemic for Schistosoma japonicum in Hunan province. The questionnaire, comprising 15 multiple-choice questions, was administered by teachers in order to collect both ethnographic and epidemiological data relevant to current S. japonicum infections. This was followed by Kato-Katz thick smear stool examinations, miracidium hatching tests, and soluble egg antigen-enzyme linked immunosorbent assays in order to validate the efficacy of the questionnaire approach. The results from a combination of all 3 procedures indicated that the overall schistosomiasis prevalence in the 3 schools was 29·9% ( 138 472 ). Six risk factors (episodes of diarrhoea, frequency of water contact, school grade attained, weakness, past history of S. japonicum infection(s), and whether a subject had been previously treated for schistosomiasis) in the questionnaire were determined by logistic regression to be highly statistically significant predictors of individual current infection. The sensitivity (93·7%), specificity (91·9%) and low cost ( c. US$ 0·6/true positive case) associated with the 6 variables model make the questionnaire approach a very useful diagnostic tool for screening marshland and lake communities at high risk for schistosomiasis in China before selective treatment with praziquantel or diagnostic follow-up. An even simpler 3 variables ‘yes/no’ model was derived from the questionnaire and found to be nearly as good at predicting individual infection (sensitivity 86·2% and specificity exceeding 97·6%) and extremely simple to use. If validated in other ecological settings in China the questionnaire, modified or as presented here, could be adopted by the national schistosomiasis control programme.