The main objective of this paper is to review the literature regarding studies of cognitive behaviour therapy (CBT) for people with first episode psychosis or early psychosis. A comprehensive search of the PsycINFO and MEDLINE databases identified twelve studies, including several randomised controlled trials and quasiexperimental studies. There were few significant differences between groups at end of treatment or at follow-up, and no differences in rates of relapse or readmission. The findings suggest, however, that CBT does have important benefits in terms of rate of recovery, improvements in certain symptoms (e.g. auditory hallucinations and hopelessness) and quality of life. Thus, there is modest support for the application of CBT for people experiencing early psychosis; however, the studies to date have difficulties with study design, theoretical underpinnings and how well the treatment targets fit with the CBT model. Future research addressing these issues is required to determine whether CBT for early psychosis may prove to be more effective than current data would suggest.