Background: Clozapine (clozaril, Mylan Products Ltd) is a first-choice treatment for people with schizophrenia who have a poor response to standard antipsychotic medication. However, a significant number of patients who trial clozapine have an inadequate response and experience persistent symptoms, called clozapine-resistant schizophrenia (CRS). There is little evidence regarding the clinical effectiveness of pharmacological or psychological interventions for this population. Objectives: To evaluate the clinical effectiveness and cost-effectiveness of cognitive-behavioural therapy (CBT) for people with CRS and to identify factors predicting outcome. Design: The Focusing on Clozapine Unresponsive Symptoms (FOCUS) trial was a parallel-group, randomised, outcome-blinded evaluation trial. Randomisation was undertaken using permuted blocks of random size via a web-based platform. Data were analysed on an intention-to-treat (ITT) basis, using random-effects regression adjusted for site, age, sex and baseline symptoms. Cost-effectiveness analyses were carried out to determine whether or not CBT was associated with a greater number of quality-adjusted life-years (QALYs) and higher costs than treatment as usual (TAU). Setting: Secondary care mental health services in five cities in the UK. Participants: People with CRS aged = 16 years, with an International Classification of Diseases, Tenth Revision (ICD-10) schizophrenia spectrum diagnoses and who are experiencing psychotic symptoms. Interventions: Individual CBT included up to 30 hours of therapy delivered over 9 months. The comparator was TAU, which included care co-ordination from secondary care mental health services. Main outcome measures: The primary outcome was the Positive and Negative Syndrome Scale (PANSS) total score at 21 months and the primary secondary outcome was PANSS total score at the end of treatment (9 months post randomisation). The health benefit measure for the economic evaluation was the QALY, estimated from the EuroQol-5 Dimensions, five-level version (EQ-5D-5L), health status measure. Service use was measured to estimate costs. Results: Participants were allocated to CBT (n = 242) or TAU (n = 245). There was no significant difference between groups on the prespecified primary outcome [PANSS total score at 21 months was 0.89 points lower in the CBT arm than in the TAU arm, 95% confidence interval (CI) -3.32 to 1.55 points; p = 0.475], although PANSS total score at the end of treatment (9 months) was significantly lower in the CBT arm (-2.40 points, 95% CI -4.79 to -0.02 points; p = 0.049). CBT was associated with a net cost of £5378 (95% CI -£13,010 to £23,766) and a net QALY gain of 0.052 (95% CI 0.003 to 0.103 QALYs) compared with TAU. The cost-effectiveness acceptability analysis indicated a low likelihood that CBT was cost-effective, in the primary and sensitivity analyses (probability < 50%). In the CBT arm, 107 participants reported at least one adverse event (AE), whereas 104 participants in the TAU arm reported at least one AE (odds ratio 1.09, 95% CI 0.81 to 1.46; p = 0.58). Conclusions: Cognitive-behavioural therapy for CRS was not superior to TAU on the primary outcome of total PANSS symptoms at 21 months, but was superior on total PANSS symptoms at 9 months (end of treatment). CBT was not found to be cost-effective in comparison with TAU. There was no suggestion that the addition of CBT to TAU caused adverse effects. Future work could investigate whether or not specific therapeutic techniques of CBT have value for some CRS individuals, how to identify those who may benefit and how to ensure that effects on symptoms can be sustained.