The ways in which differences between expected grades and those actually received affect a student's academic expectations, actual performance, attribution of success or failure, and self-involvement were measured. Using the assumption that an individual is usually motivated to maintain or increase his self-esteem, 4 hypotheses concerning academic performance were constructed, and instruments were developed to collect data from a random selection of 220 freshmen (110 males, 110 females) entering Bucknell University in Fall 1965. Data were collected in the first week of the first (Time 1) and second (Time 2) semester to measure the students' level of aspiration (expected GPA), self-involvement, and self-esteem. Success or failure was determined by comparing discrepancies between levels of aspiration at Time 1 and actual GPAs at Time 2. A higher GPA than anticipated was interpreted as success, and a lower as failure. The data indicated that there seems to be no relationship between success and failure and changes in self-esteem; that students' expectations regarding future academic performance is related to current academic success or failure; and that for women, self-involvement in the student role increases with good performance and decreases when they perform poorly. Implications of the findings for the US educational system are discussed. (WM)