Following a review of the relevant literature a Cognitive Behavioural treatment protocol for the prevention of relapse in schizophrenia is presented. This treatment protocol is investigated in a 12-month non-blind randomised controlled trial comparing Cognitive Behavioural Therapy and Treatment as Usual (CBT + TAU) versus Treatment as Usual (TAU) alone. Three studies of treatment outcome are described: relapse and admission, remission and social functioning, and psychological distress. 144 participants with a DSM-IV Schizophrenia spectrum disorder were randomised to receive either CBT + TAU (n = 72) or TAU alone (n = 72). 11 participants dropped out (6 from CBT + TAU, 5 from TAU alone) leaving a completers sample of 133. Participants were assessed at entry, 12-weeks, 26-weeks, and 52 weeks. CBT was delivered over two stages: a 5-session engagement phase which was provided between entry and 12-weeks, and a targeted CBT phase which was delivered on the appearance of early signs of relapse. Over 12-months CBT + TAU was associated with significant reductions in relapse and admission rate. The clinical significance of the reduced relapse and admission rate amongst the CBT + TAU group was investigated. First, receipt of CBT + TAU was associated with improved rates of remission over 12-months. Second, clinically significant improvements in social functioning were investigated. Again, receipt of CBT + TAU was associated with clinically significant improvements in prosocial activities. However, receipt of CBT + TAU was not associated with improvements in psychological distress over 12-months. The theory underpinning the cognitive behavioural treatment protocol predicted that negative appraisals of self and psychosis represent a cognitive vulnerability to relapse. This hypothesis was investigated during the present 2 Abstract study. After controlling for clinical, treatment and demographic variables, negative appraisals of self and entrapment in psychosis were associated with increased vulnerability to relapse, whilst negative appraisals of self were associated with reduced duration to relapse. Finally, an explorative study of changes in negative appraisals of psychosis and self over time, which were associated with relapsers versus non-relapsers from the TAU alone group, was conducted. This study found a strong association between the experience of relapse, increasing negative appraisals of psychosis and self, and the development of psychological co-morbidity in schizophrenia. Results of treatment outcome and theoretical analyses are discussed in terms of their relevance to the further development of psychological models and treatments for psychosis.