Abstract: Objective To carry out a theoretical research on the rule that a mathematical model may play on predicting and evaluating the control effect of schistosomiasis japonica. Methods Barbour＇s two-host model was used to predict and evaluate the effect of different control measures by computer simulation. Pilot samples in two villages of Shanghai suburb in 1950s were applied for the analysis. Results When the prevalence was high, synchronous chemotherapy for human and cattle populations quickly reduced the indices of the infection. Mollusciciding provided positive impact on the effect of chemotherapy. Added with the environmental measures for snail control，the basic reproductive rate (BRR) and equilibrium prevalence in human and bovines were sustainably reduced and even reached an interruption of transmission. The effect of chemotherapy could be consolidated by the anti-fecundity vaccine for bovines. Satisfied control effect could also be obtained by chemotherapy in human and bovines combined with behaviour intervention for human and vaccination for bovines without snail control. In areas with lower levels of transmission velocity，BRR and prevalence，the effect of various interventions was better than that obtained in areas with higher levels of the above three infection indices，and the disease control could be easier. Conclusion The Barbour＇s mathematical model can be used to roughly predict and evaluate the effects of schistosomiasis control measures.

Key words: Schistosomiasis japonica, Mathematical model, Evaluation, Control effect