BackgroundCannabis abuse in psychotic patients is associated with rehospitalizations, reduced adherence and increased symptom severity. Previous psychosocial interventions have been ineffective in cannabis use, possibly because of low sample sizes and short interventions. We investigated whether adding CapOpus to treatment as usual (TAU) reduces cannabis use in patients with cannabis use disorder and psychosis.MethodA total of 103 patients with psychosis and cannabis use disorder were centrally randomized to 6 months of CapOpus plus TAU (n = 52) or TAU (n = 51). CapOpus consisted mainly of motivational interviewing and cognitive behaviour therapy (CBT). TAU was targeted primarily at the psychotic disorder. The primary outcome was self-reported days with cannabis use in the preceding month.ResultsPre-randomization cannabis use frequency was 14.9 [95% confidence interval (CI) 12.7–17.1] days/month. Post-treatment, the ratio of days/month with cannabis use in CapOpus versus TAU was 0.76 (95% CI 0.38–1.50) (p = 0.42), and 0.80 (95% CI 0.21–3.10) (p = 0.75) at the 4-month follow-up. From 46.4 (95% CI 36.4–56.3) monthly joints pre-randomization, consumption fell to 27.3 (95% CI 12.6–41.9) joints in CapOpus and 48.2 (95% CI 31.8–64.6) in TAU (p = 0.06). Follow-up amounts were 28.4 (95% CI 13.5–43.2) and 41.6 (95% CI 25.2–58.0) joints (p = 0.23). Several subgroup analyses suggested benefits of CapOpus.ConclusionsCapOpus did not reduce the frequency, but possibly the amount, of cannabis use. This is similar to the findings of previous trials in this population. Implementation of CapOpus-type interventions is thus not warranted at present but subgroup analyses call for further trials.