Cognitive therapy is now widely recommended for the treatment of schizophrenia. However, little is known about factors predicting good outcome. We are describing separate sub-analyses of two randomized controlled trials of Cognitive Behaviour Therapy for schizophrenia. In one trial expert therapists provided therapy while in the second trial brief therapy was provided by trained nurses. In both trials psychopathology was assessed using the Comprehensive Psychopathological Rating Scale (CPRS) and the Health of the Nation Outcome Scale (HoNOS). Insight, delusions, hallucinations and negative and positive symptoms were also measured. In brief therapy trial higher levels of insight and high CPRS global impression score predicted good outcome in the CBT group. Analyses using similar variables did not reveal any specific predictors for good outcome in the second trial.