The general hypothesis of this study was that college students, according to their perception of the amount of learning achieved in a class, will judge the quality of the interpersonal relationships between them and their teachers differently. Specifically, the hypotheses tested were that (1) students in "learn-most" classes would perceive themselves as receiving from their teacher more regard as individuals, more empathetic understanding, more acceptance of themselves as unique individuals, and more feelings of being totally themselves, than would students in "learn-least" classes; and (2) There would be differences between the "learn-most" and "learn-least" students on the basis of sex, grade received, subject, level of course, and student year. The instrument employed to obtain a measure of interpersonal relationships was the Barrett-Lennard Relationship Inventory (1962). One hundred eighteen college students, randomly assigned to two groups, completed the inventory. One group responded with their perceptions of the teacher from whom they learned the most, and the other group responded with their perceptions of the teacher from whom they learned the least. Findings tended to substantiate the general hypothesis. "Learn most" students judged the quality of their interpersonal relationships with their teachers more positively than did "learn least" students. The second hypothesis was not proved. The implications of the findings are that teacher education programs should emphasize the ability to develop positive interpersonal relationships with students rather than methods and materials. (JD)