BACKGROUND AND OBJECTIVES: Cognitive behaviour therapy (CBT) for psychosis has been shown to be effective, but there are recent suggestions that it is less efficacious than initially thought. Metacognitive therapy (MCT), which focuses on metacognitive mechanisms, has led to positive results in other disorders, but has yet to be evaluated in people with schizophrenia spectrum diagnoses. This study evaluates the feasibility of MCT for people with psychotic disorders. METHODS: Ten participants with schizophrenia spectrum disorders received up to 12 sessions of MCT in an open trial. Outcomes included psychiatric symptoms measured using the PANSS, at baseline, 9 months (end of treatment) and at 12 months (follow-up), as well as dimensions of hallucinations and delusions, emotional dysfunction, self-rated recovery, social functioning and metacognitive beliefs. RESULTS: T-tests and Wilcoxon's signed ranks tests revealed significant beneficial effects on several outcomes at end-of-treatment and follow-up. Cohen's d effect sizes were moderate to large (for PANSS total, d = 1.0 at end of treatment; d = 0.95 at follow-up). A response rate analysis found 50% and 40% of participants achieved at least a 25% reduction in PANSS total scores by end of therapy and follow-up, respectively. Exploratory analyses revealed that metacognitive beliefs significantly changed over treatment and follow-up periods. LIMITATIONS: This study had no control group and was not randomised; therefore, it is likely that effect sizes were inflated. CONCLUSIONS: This study provides preliminary evidence that MCT is a feasible treatment for people with psychosis. An adequately powered randomised controlled trial is warranted.