BACKGROUND: In view of the plethora of different community-based mental health services, there is a clear need for a classification based on service components rather than labels. Moreover, the sustainability of experimental services beyond their research studies is rarely reported. METHODS: As part of a systematic review of home treatment for mental health problems, authors of all included studies were followed up for data on service components and sustainability. Associations between components were explored. RESULTS: There was evidence of a core group of components co-occurring in home treatment services: regularly visiting at home, taking responsibility for health and social care, having smaller caseloads, multidisciplinary teams, integrated psychiatrists and a high proportion of contacts at home. Fifty-four per cent of services no longer existed, of which almost half had ended by the study's publication date. There was a significant association between sustainability and the study's hospitalisation outcome. CONCLUSIONS: Some of the related service components presented here were associated with reducing hospitalisation. This group cannot, however, be used to provide a new taxonomy of services. It is imperative that future studies prospectively record and report service components to enable better classification. It is of concern that policy is currently predicated on research findings regardless of whether or not the experimental service was sustainable.