Background: Historically, it has been difficult to demonstrate an impact of training in psychological interventions for people with psychosis on routine practice and on patient outcomes. A recent pilot evaluation suggested that postgraduate training in Cognitive Behavioural Therapy for Psychosis (CBTp) increased the delivery of competent therapy in routine services. In this study, we evaluated clinical outcomes for patients receiving therapy from therapists who successfully completed training, and their association with ratings of therapist competence and therapy content. Aims: To characterize the therapy delivered during training and to inform both a calculation of effect size for its clinical impact, and the development of competence benchmarks to ensure that training standards are sufficient to deliver clinical improvement. Method: Paired patient-reported outcome measures (PROMS) were extracted from anonymized therapy case reports, and were matched with therapy ratings for each therapist. Results: Twenty clients received a course of competent therapy, including a high frequency of active therapy techniques, from nine therapists. Pre–post effect size for change in psychotic symptoms was large (d = 1.0) and for affect, medium (d = 0.6), but improved outcomes were not associated with therapist competence or therapy content. Conclusions: Therapists trained to research trial standards of competence achieved excellent clinical outcomes. Therapy effect sizes suggest that training costs may be offset by clinical benefit. Larger, methodologically stringent evaluations of training are now required. Future research should assess the necessary and sufficient training required to achieve real-world clinical effectiveness, and the cost-effectiveness of training.