The CRT model of treating acute mental health crises outside inpatient wards has been implemented in several Western countries in the past decade. In addition to this adoption, the CRT model has been implemented as part of national policies in Norway and the UK. This makes research on the CRT model highly relevant. Our study examines Norwegian CRTs and their patients, treatment outcome and pattern of admissions from the CRTs to inpatient wards. The sample consisted of 680 patients and 62 staff members of eight Norwegian CRTs. The data were collected in 2005 and 2006. A registration form was developed to record information about the patients from admission to discharge, including socio-demographic, and clinical data, the content of treatments and the pattern of admissions from the CRTs. The Community Program Practice Scale (CPPS) was completed by each CRT clinician and a questionnaire on how the CRTs were organized and operated was completed by the team leaders of each CRT. We found that the Norwegian CRTs operate without gate-keeping function for acute inpatient wards, without 24/7 operating hours, with 40 per cent of patients waiting more than 24 hours for treatment and with patients who were not considered for hospital admission being treated. The CRTs worked more with depression and suicidal crises than with psychoses. Compared to the intentions of the CRT model, the CRTs provided less intensive and less out-of-office care. The odds of being admitted to in-patient wards were significant lower for those patients treated by a CRT operating extended opening hours compared to CRTs operating in office hours only. In addition, patients with psychotic symptoms, with concrete suicidal plans or self-injury but no death intention, and with a prior history of admissions were more likely to be admitted. These findings indicate CRTs in Norway operate in a way that departs from the CRT model, and that there are reason to believe that the CRTs do not completely fulfil their intended role in the mental health system.