A self-talk package was used to improve performance of compulsory figures by prenovice- and novice-level figure skaters. The study included ongoing objective behavioral assessment across practices of the figure skating performance as well as the extent to which the skaters actually utilized the self-talk. A multi-element design with multiple baseline replications across 4 participants demonstrated that improvements were due to the treatment. Self-report follow-up at 1 year indicated that the participants continued to utilize the selftalk during practices and that they believed that it enhanced their test and/or competitive performance. The results support the view that planned self-talk can aid skill acquisition. Results are conceptualized in terms of rule-governed control over behavior, which may provide a useful framework for enabling sport psychologists to increase the efficacy of self-talk interventions.