An investigation was undertaken of end-of-course reactions of students who experienced varying numbers of test-taking failures during an introductory psychology course presented under two different formats: a mastery structure (multiple study/test options and absolute grading standards) and a conventional structure (one try on each midterm with relative standards). The subjects were 74 students who experienced subjective failure on both of two midterms under conventional circumstances and 145 students who failed two, three, or four times under conditions of repeated test opportunities. To assess the impact of accumulating failures, students responded to a final course evaluation form consisting of 28 questions. Subsequent factor analysis reduced these responses to six factors. Because students took more tests under the mastery condition, more subjective failure incidents were reported. Yet despite the greater absolute number of failure experiences, there was no greater deterioration of feelings of personal control, achievement orientation, or sense of enjoyment. Overall, the mastery students expressed significantly greater confidence and aspirations and assessed the system as more fair. Path analysis identified the cause of the continuing positive motivation as increased objective performance, which itself depends on the multiple study/test option. It is concluded that as long as students show increasing objective performance, short-lived failures along the way play little part in determining overall course evaluation. Behavioral instruction appears to be a beneficial system for all students, especially for slow learners. (Author/SW)