Research suggests that depressed patients' beliefs about the causes of their depression affect the course and outcomes of treatment. However, changes in such beliefs during treatment have not been investigated. Before and after treatment, participants in a randomized control trial comparing cognitive therapy (CT) with antidepressant medication (ADM) in the treatment of depression were asked about their beliefs concerning the reasons for their depression. The reason clusters of interest were: Characterological, conceptually related to CT; and Biological, conceptually related to ADM. Contrary to predictions, endorsement of treatment-relevant reasons did not increase after a successful treatment. Rather, endorsement of treatment-irrelevant reasons decreased significantly. Thus, successful treatment may preserve people's beliefs in causes of depression most closely related to the treatment, and diminish beliefs in unrelated causes. © 2007 Springer Science+Business Media, LLC.