Schistosomiasis is the second widespread tropical disease that affects the health of over 240 million people of 78 countries. Questionnaires have been commonly used to diagnose schistosomiasis, while no meta-analysis of their efficacy had been reported previously. This meta-analysis was conducted to assess their diagnostic accuracy of schistosomiasis. Studies published prior to December 1, 2014, that had used questionnaires as a diagnostic tool were searched in PubMed, Medline, EMBASE, and China National Knowledge Infrastructure (CNKI) database. A total of 32 studies with 72,812 cases were identified for the meta-analysis. The best diagnostic odds ratio (DOR) was obtained from Schistosoma haematobium (67.68, 95 % confidence interval (CI) = 31.48 to 145.54), followed by Schistosoma japonicum (11.74, 95 % CI = 4.59 to 30.05) then Schistosoma mansoni (2.98, 95 % CI = 1.95 to 4.54). Pooled sensitivity and specificity were respectively 0.82, 0.88, and 0.46, and 0.59, 0.86, and 0.81 for S. japonicum, S. haematobium, and S. mansoni. The multivariable subgroup analyses showed that the questionnaires which had been pretested and standardized had better diagnostic performance. The result suggested that questionnaires can be used to diagnose schistosomiasis with moderate sensitivity and specificity and the questionnaires for diagnosing S. haematobium performed best.