DescriptionThe use of paradoxical approaches to psychotherapy has an extensive history dating to Adler. The practice of paradoxical therapy has been hindered by the lack of a testable theoretical framework. Likewise, little experimental evidence exists for the efficacy of paradoxical techniques. This dissertation examines one paradoxical therapy technique, paradoxical intention (PI), from a social-learning theory perspective. The growing experimental literature on PI is critically reviewed with regard to the methodological adequacy of the studies, and the evidence for the efficacy of PI. A testable theoretical explanation of the action of PI is proposed drawing upon two segments of the experimental-clinical literature: l) the effects of uncertainty of aversive stimulation on behavior and arousal, and 2) Bandura's self-efficacy theory. Finally, a program of research is proposed to test aspects of the proposed theory as well as to answer other questions about PI. It is concluded that PI is a potentially facilitative adjunct to exposure methods of anxiety and arousal reduction, but that, due to problems of confounding of PI and exposure treatments in the experimental literature, it is unclear whether PI has any specific therapeutic effects beyond facilitation of exposure.