Objectives. Studies investigating the efficacy of cognitive-behavioural treatment for post-stroke depression have produced contradictory results (Flannaghan, 2000; Lincoln, Flannaghan, Sutcliffe, & Rother, 1997). This study aimed to assess whether depressed stroke patients experienced more negative and less positive cognitions than non-depressed stroke patients in order to ascertain whether post-stroke depression is suitable for cognitive-behavioural therapy. Design. A questionnaire-based study compared stroke patients who were assessed as depressed or not depressed. Methods. A Cognitions scale was developed from transcripts of cognitive-behavioural therapy sessions with depressed stroke patients. A total of 50 stroke patients, the majority of whom were still in hospital, were assessed on the Beck and Wakefield Depression Inventories and the Cognitions scale. Results. The prevalence rate for depression was similar to that found in previous studies. The internal reliability and construct validity of the scale were found to be acceptable. Significantly more negative cognitions and less positive cognitions were found in stroke patients who were depressed as compared with stroke patients who were not depressed. Conclusions. The study suggests that post-stroke depression does not differ qualitatively from general depression and that general theories and thus treatments for depression may be valid within this population.