Panic attacks are, by definition, frightening experiences. Recent research has attempted to develop efficacious treatments for individuals suffering from these sudden episodes of fear. The current investigation sought to examine the efficacy of a new cognitive-behavioral treatment, Guided lmaginal Coping. This treatment aims to improve the coping repertoires of panic sufferers. Pilot results with this treatment suggest that it is an effective treatment. The current study sought to extend these findings by evaluating the process of change over time. The second purpose of the current investigation was to examine whether the concept of self-efficacy could help explain the process of change over time. Nineteen subjects diagnosed with Panic Disorder or Agoraphobia with panic attacks were assigned to one of two treatment groups: Guided lmaginal Coping or Panic Education, a nondirective treatment. Subjects in each group received six individual and four group therapy sessions. In addition to treatment, each subject met with a trained evaluator before treatment began, the five weeks of treatment, posttreatment, and one and two month follow-up periods. During these evaluations, subjects participated in a taped anxiety induction procedure and completed questionnaires concerning their degree of self-efficacy, level of panic symptoms, catastrophic thoughts, avoidance, and frequency of coping. Results indicated that subjects in both treatment groups had significant reductions in symptoms, catastrophic thoughts, and number of attacks. Guided lmaginal Coping subjects demonstrated reduction in avoidance and showed continued improvement in symptom reduction during the follow-up period, Both groups showed improved coping abilities. Self-efflcacy and rated coping were significantly related at all evaluation periods. Cross-lagged panel analyses revealed that self-efficacy appeared to lead to changes in catastrophic thoughts whereas self-efficacy and symptom level appeared mutually causal. Self-efficacy and avoidance did not appear consistently related nor did self-efficacy and coping frequency. Rated coping effectiveness appeared to improve in both groups although coping frequency did not. Results are discussed related to treatment outcome and to panic not being as refractory a problem as presumed. Implications are discussed in terms of models of panic and the significance of self-efficacy and catastrophic cognitions to these models and to the amelioration of panic.