Based on a behavioral analysis of compliance with breast self-examination (BSE), this study replicates a previously successful stimulus control intervention and tests the effectiveness of two types of reward for increasing compliance: external reward and self-reward. Patient volunteers (N = 153) were randomly assigned to experimental conditions and followed for 1 year, equally divided into experimental and postexperimental periods. Compliance was measured by the monthly return of BSE records, with self-reports used as a secondary measure. The results indicate that external reward was most effective; self-reward was used by only half the participants but was effective when used. In the postexperimental period, all rates of record return declined, especially in the external reward group. A cycle-by-cycle analysis indicates that the decline was gradual except for a sharp drop when the external reward was withdrawn. The results are discussed in terms of the acceptability of both interventions and contingencies.