This study examined how college students' test anxiety related to previous testing experiences and self-efficacy. Participants were 29 undergraduate college students enrolled at a Midwestern university. Instruments used to measure the criterion and predictor variables were: Test Anxiety Scale, College Academic Self-efficacy Scale, Generalized Self-Efficacy Scale, and Previous Testing Experience Scale. Two separate simultaneous multiple regression analyses were performed to answer the research questions. Results indicate that the regression equation of test anxiety using previous testing experiences and academic self-efficacy as predictors was significant, and 40% of the variability in test anxiety was predictable on the basis of these two predictor variables. On the other hand, using general efficacy instead of academic efficacy explained less variability, and the general efficacy variable was found to be not significant. Correlational analysis demonstrated that test anxiety is positively associated with previous testing experience and negatively related with academic self-efficacy. The findings supported the previous research conducted in this field. (Contains 3 tables and 18 references.) (Author/SLD)