Background: frail older people often require tailored rehabilitation in order to remain at home, especially following a period of hospitalisation. Restorative care services aim to enhance an older person's ability to remain improve physical functioning, either at home or in residential care but evidence of their effectiveness is limited. Objective: to evaluate the effectiveness of a restorative care service on institutional-free survival and health outcomes in frail older people referred for needs assessment in New Zealand. Methods: a randomised controlled trial of restorative care or usual care in 105 older people at risk of permanent residential who were follow-up over 24 months. The restorative care service was delivered in short-stay residential care facilities and at participants' residences with the aim of reducing the requirement for permanent residential care. It included a comprehensive geriatric assessment and care plan developed and delivered, initially by a multi-disciplinary team and subsequently by home care assistants. Results: compared with usual care, there was a non-significant absolute risk reduction of 14.3% for death or permanent residential care (8.8% for residential care and 7.2% for death alone) for the restorative care approach. There was no difference in levels of burden among caregivers. Conclusions: restorative care models that utilise case management and multi-disciplinary care may positively impact on institutional-free survival for frail older people without adversely impacting on the health of caregivers.