Although social-learning theory remains among the dominant perspectives in micro-level research concerning the causes of delinquency, research has yet to provide an adequate test of its social-reinforcement component using the methods required by the logic of operant psychology. The present article discusses the limits of existing attempts to test the social-reinforcement hypothesis, offers a new approach for testing it, and describes the use of panel data to provide such a test. In particular, the author examined (1) whether delinquency increases the amount of time a perpetrator's peers choose to spend with him or her, (2) whether such attention serves as a direct reinforcement prompting further delinquency from the perpetrator, and (3) whether such attention serves as a vicarious reinforcement prompting delinquency from audience members in proportion to their desire for informal socializing among peers. The results suggest support for vicarious, but not direct, social reinforcement. (Contains 5 tables and 2 figures.)