The effects of two types of contingency contracts and self-pacing on student performance in a personalized system of instruction (PSI) class were examined. Forty-seven college students in a principles of educational psychology class were randomly assigned to either a short contract, long contract or self-pacing group. One group of students signed a short contract requiring mastery (90 percent) of the first unit within 3 class days, while another group signed a longer contract. A final group of students were self-pacing from the beginning of the course, and no contract was signed. Students in the three groups were also assigned to peer tutors. Analysis of covariance (with grade point average and student classification as covariates) yielded significant differences only on the final examination among the three groups. Multiple t-test comparisons indicated that both the self-pacing group and the short contract group performed significantly higher than did the long contract group on the final examination. (Author/SW)