The classroom test performance of three groups, self-monitoring (SM), a combination of self-monitoring, self-reinforcement and self-punishment (SM+C), and controls, were compared in independent replications in a chemistry (n=149) and a calculus (n=80) class. In chemistry, but not calculus, the experimental subjects outperformed controls. It was predicted that SM+C would outperform SM students and would also report more study time, but these hypotheses were not confirmed. Reported study time differed between the two classes and the course difficulty level also appeared to differ. Course difficulty may be affecting the successful use of self-control techniques. (Author)