The authors examined the effects of randomized and delayed credit on the percentage of college students participating at four pre-defined levels on each class discussion day: non-participation, credit-level participation, frequent participation, and dominant participation. Although the same amount of participation credit was available to students in three relatively large course sections, there were some procedural differences across sections as to when and by whom credit units and credit days within those units were selected. Findings indicated that student perceptions of uncertainty and student choice regarding credit contingencies help explain both the immediate and continuing effects of credit contingencies on students' participation patterns. (Contains 1 table and 1 figure.)