Aim: The finding by Jacobson and colleagues that there was no difference in treatment outcome for individuals treated with a complete cognitive therapy (CT) package and those receiving behavioural activation raised serious concerns about the mediational role of cognition in depression. The present study was designed to test whether the interaction between coping and cognition predicts changes in depressive symptomatology. Method: The sample ( N =30) was derived from the full CT condition of the study conducted by Jacobson and colleagues. Therapy transcripts from the third session were rated to assess cognitive errors and coping. Change in depression scores from pre‐treatment to post‐treatment was computed. Findings: Analyses demonstrated that engagement in self‐reliance and not engaging in escape coping predicted greater decreases in levels of depression after CT. Implications for practice: These findings suggest that engaging in specific types of adaptive coping strategies and avoiding maladaptive coping may be important in predicting decreases in depression for individuals who have high levels of certain cognitive errors.