The Elaboration Likelihood Model (ELM; Petty & Cacioppo, 1981a) of attitude change provides a framework for understanding the differential impact of central characteristics (e.g., intervention quality) and peripheral features (e.g., source credibility) of communication in clinical contexts. This study reports an application of the model to pretreatment attitudes regarding a specific therapy (i.e., cognitive restructuring) for treating eating disorders. Results of the study generally supported the utility of the ELM in this analogue clinical context. Although intervention quality was the primary determinant of attitude change, it interacted with source credibility to predict measures of behavioral compliance in a way that is consistent with the model. Implications and limitations of these findings are discussed.