A Partial Least Squares Path Analysis technique was used to test the effect of students' prior experience with computers, statistical self-efficacy, and computer anxiety on their achievement in an introductory statistics course. Computer Anxiety Rating Scale and Current Statistics Self-Efficacy Scale were administered to a sample of 64 first-year university undergraduates (35 males and 29 females) enrolled in an introductory statistics course in a Faculty of Education. Achievement scores were obtained from students' records. Results of the study revealed that statistical self-efficacy was the most important predictor of students' achievement in statistics, followed by prior experience with computers and finally computer anxiety. In addition, statistical self-efficacy and prior experience with computers had an indirect effect on achievement in statistics through their effect on computer anxiety. Implications of these findings for teaching and learning statistics are discussed. (Contains 3 figures and 3 tables.)