BACKGROUND: The effects of hospital-based rehabilitation including weekly supportive psychodynamic therapy compared with specialized assertive intervention and standard treatment has not previously been investigated in first-episode psychosis. The aim of the study was to examine long-term effect on use of institutional care of different intensive interventions for patients with first-episode schizophrenia spectrum disorder on use of psychiatric bed days and days in supported housing.MethodA total of 94 severely ill patients with first-episode schizophrenia spectrum disorders were included in a special part of the Copenhagen OPUS trial and randomized to either the specialized assertive intervention program (OPUS), standard treatment or hospital-based rehabilitation. RESULTS: It was a stable pattern that patients randomized to hospital-based rehabilitation spent more days in psychiatric wards and in supported housing throughout the 5-year follow-up period compared with the two other groups. Patients in OPUS treatment spent significantly fewer days in psychiatric wards and supported housing in the first 3 years compared with patients in hospital-based rehabilitation. Due to attrition and small sample size, differences in level of psychotic and negative symptoms at 5-year follow-up could not be evaluated. CONCLUSIONS: The study indicates that hospital-based rehabilitation together with weekly supportive psychodynamic therapy was associated with a continued increased use of psychiatric bed days and days in supported housing. The data cannot justify using hospital-based rehabilitation in first-episode psychosis.