Investigated the treatment effects of three social influence variables frequently implicated in psychotherapy placebos. Socially anxious male Ss participated in an experimental treatment for reducing dating anxiety. Ss were either given or not given specific suggestions for decreasing social anxiety, placed in conditions of high or low social demand, and received feedback indicating either high or moderate success on 'the therapy task. Results support the importance of social influence variables in therapeutic change. Contingent success had its greatest impact on personal attributes; suggestion affected skill behaviors; and social demand effects were found in the selfevaluation of heterosocial performance. Various social influences appear to mediate change differently and do not exert the generic effects commonly assumed to be characteristic of therapy placebos. Implications for outcome research are discussed.Variables associated with placebo effects have received much attention as alternative explanations for what causes therapy to work (Kazdin, 1979), but relatively little is known about how they mediate change (Bootzin & Lick, 1979). Conceptual confusion over the placebo construct has impeded meaningful investigation on the treatment influences of such variables as suggestion, social demand, and persuasion (Critelli & Neumann, 1984). Wilkins (1986) noted that it is counterproductive for psychotherapy research to dismiss these variables as nonspecific artifacts because they actually constitute and should be studied as legitimate social psychological models of change. The particular social psychological variables implicated in placebo procedures will be referred to as social influence factors because they derive their impact primarily from the social processes between therapist and client rather than from any learning experiences or skill development resulting from treatment.Research on placebo procedures has focused on the credibility of the treatment rationale (Kazdin & Krouse, 1983), the style by which the treatment is administered (e.g., therapist directiveness or attention), and the type of procedures used to implement the therapeutic tasks (e.g., homework assignments; Jacobson & Baucom, 1977). However, several other social influences may play prominent roles in the efficacy of psychotherapeutic interventions.This article is based on Nolan W. S. Zane's doctoral dissertation submitted to the Graduate School of the University of Washington in partial fulfillment of the doctoral degree. Parts of this research were presented at the annual meeting of the Western Psychological