The present study was conducted to investigate the specific and differential effects of test anxiety and intelligence on academic achievement in four subject areas. Manipulated in a 3 X 3 factorial design, three levels of test anxiety and three levels of intelligence were used as independent variables. 22 Ss were randomly assigned to each treatment condition. "Thus, 198 Ss were studied. An analysis of variance for each subject area indicated significant main effect of intelligence only. However, interaction effects between anxiety and intelligence were significant in all subject areas. In summary, the results showed that there was a different optimum level of test anxiety which might facilitate achievement in different subject areas at different ability levels.