Around 60 000 people in England live in mental health supported accommodation. There are three main types: residential care, supported housing and floating outreach. Supported housing and floating outreach aim to support service users in moving on to more independent accommodation within 2 years, but there has been little research investigating their effectiveness.

A 30-month prospective cohort study investigating outcomes for users of mental health supported accommodation.

We used random sampling, accounting for relevant geographical variation factors, to recruit 87 services (22 residential care, 35 supported housing and 30 floating outreach) and 619 service users (residential care 159, supported housing 251, floating outreach 209) across England. We contacted services every 3 months to investigate the proportion of service users who successfully moved on to more independent accommodation. Multilevel modelling was used to estimate how much of the outcome and cost variations were due to service type and quality, after accounting for service-user characteristics.

Overall 243/586 participants successfully moved on (residential care 15/146, supported housing 96/244, floating outreach 132/196). This was most likely for floating outreach service users (versus residential care: odds ratio 7.96, 95% CI 2.92–21.69, P < 0.001; versus supported housing: odds ratio 2.74, 95% CI 1.01–7.41, P < 0.001) and was associated with reduced costs of care and two aspects of service quality: promotion of human rights and recovery-based practice.

Most people do not move on from supported accommodation within the expected time frame. Greater focus on human rights and recovery-based practice may increase service effectiveness.