Objective. The Suitability for Short-Term Cognitive Therapy (SSCT) rating procedure has predicted outcome in depressed and anxious patients. This study examines its relevance in assessing patients with psychosis. Method. Outpatients with psychosis (n=56), depression (n=93), and anxiety (n=264) received cognitive- behavioral therapy in a university hospital teaching unit (mean number of sessions=16, SD=11). Demographic, clinical, and suitability variables were assessed as potential predictors of dropout and success as measured by the Reliable Change Index. Results. Despite lower suitability scores in the psychosis group, dropout and success rates were similar across groups, although the magnitude of symptom reduction was less in the psychosis group. Across diagnoses, dropout was predicted by unemployment and by reluctance to take personal responsibility for change. In the psychosis group only, dropout was predicted by hostility. Success of completed therapy was predicted by higher baseline agoraphobic anxiety and “responsibility for change” scores. Conclusion. Attention to hostility early in therapy may reduce dropout in psychotic patients. Fostering acceptance of responsibility for change may improve both treatment retention and success across diagnoses. Agoraphobic fear is associated with success, possibly reflecting the effectiveness of behavioral interventions in psychosis and anxiety alike.