In sizeable discussion-based college classrooms, achieving any approximation of balanced student participation is difficult. More common is a pattern that develops wherein a small percentage of the class dominates discussion and a larger percentage rarely or never participates. Thus, the purpose of this study was to find ways to balance the amount of discussion across students without diminishing the relevance of discussion. Consequently, this study evaluated the efficacy of crediting participation and requiring students to self-record their daily participation. Students (N = 160) in three sections of an undergraduate educational psychology course self-recorded their comments on specially designed record cards and received credit for participation during selected phases of the study. Additionally, an observer kept track of each class discussion by coding the quantity of each student’s daily participation. Relevance and type of student responses were assessed as ancillary dependent measures, also recorded by the observer. Credit decreased the percentage of both non-participants and dominant participants, thus balancing participation across students. Self-recording had a minimal effect on participation. Neither credit nor self-recording altered relevance or type of student comments. Few overall instances of non-relevant student commenting indicated that the construct was too narrowly defined, which provides direction for future attempts to assess quality of student participation. Because these findings resulted from comparisons within and between three sections of the course, instructor behavior was also monitored daily. A secondary observer’s records revealed that instructor behaviors (i.e., type and number of questions asked and feedback given) did not inflate or diminish the effect of treatment conditions. A 50-item survey assessed student perceptions of participation at the beginning of the course and was found to significantly predict student participation. Through a series of four principal components analyses, I extracted three specific factors. Logistic regression analyses showed that the primary factor, History and Confidence regarding Participation, differentiated high and low participants as well as the total survey and the three combined factors. This factor best predicted membership in a low participant group in the non-credit units and membership in a high participant group in the credit units.