Three reports in this issue of (Derby, Fisher, & Piazza, 1996; Fisher, Grace, & Murphy, 1996; Smith, Lerman, & Iwata, 1996) extend the results of a small group of studies on self-restraint, a curious phenomenon that has been observed as a correlate of self-injurious behavior (SIB). Self-restraint appears to be a behavior that is unique to individuals who engage in SIB. That is, we have not found any published reports of individuals who self-restrain but who do not engage in SIB, or of individuals who self-restrain in conjunction with other behavior problems (e.g., aggression). This raises some interesting questions about the relationship between self-restraint and SIB: Are both responses maintained by the same reinforcement contingency? Are the responses functionally independent but related through some historical process (e.g., correlated but different contingencies of reinforcement)? Does access to one response (or its termination) serve as reinforcement for the other? In reviewing some of the research on self-restraint, Smith, Iwata, Vollmer, and Pace (1992) found tentative support for each of these three hypotheses, and their data suggested that, as is the case with SIB, self-restraint may be influenced through several mechanisms of control. Each of the studies in the present series demonstrated a clear functional relationship between SIB and self-restraint by focusing on a single operant mechanism (e.g., assessing whether contingent access to self-restraint serves as reinforcement for SIB). In what follows, we will expand on these relationships,integrate findings from previous and current studies, and suggest directions for future research.