Although computer-assisted instruction (CAI) favorably affects academic performance and attitudes toward instruction, most CAI programs target mastery but not fluency. This study used a within-subjects, crossover design to compare the effects of CAI that included fluency training with traditional study (TS) conditions on test performance and attitudes toward CAI. The CAI condition consisted of 2 forms of CAI. The results showed that the mean test score in the CAI condition was about 7 percentage points higher than the mean score in the TS condition. Also, attitude scores favored the CAI condition. These data suggest that CAI procedures that use elements of both mastery learning and fluency training may help to improve academic performance and attitudes.