Purpose -The purpose of this paper is to give an account by a rehabilitation psychiatrist of many years of involvement in shared housing for former long-stay hospital residents and other iong-term mental health service users. Design/methodology/approach -The paper offers a personal view based on developments in one iocality of East London, blending case study narrative with cited eariier pubiished papers that confirm and/or give greater detaii on specific aspects of the experience gained. Findings -Long-term mentai heaith service users, given the opportunity via shared housing to develop more lasting and naturai reiationships, proved to have social and interactive capacity that was obscured by both the conditions of hospital wards and by isoiafing accommodation "in the community". The experience provides iessons for a better understanding of patients' problems. Practical implications -Shared housing is often seen as simpiy a poor substitute for independent iiving; but the experience recounted here suggests that shared iiving may have particular value for some dient groups. Originality/value -The housing developments and the associated research remain unusual and of potential value forpianners of long-term services.Mental health services advance and regress in unexpected and puzzling ways. For this writer, for many year a practicing psychiatrist, three decades of engagement in housing for long-term patients was an enlightening privilege. In this paper, I propose to look back over the history of the development of such shared and supported housing, and the challenge they present to conventional thinking, with greater recognition of the social side of medical care.It may be surprising to discover that plans for shared houses to supplement asylum wards had been created, but then abandoned, 150 years ago, even as these institutions became intolerably overcrowded. Yet "cheerful and home-like" houses were developed from offices in the grounds of the Devon County Asylum by Bucknill, its Physician Superintendent, who set out plans for them to replace up to half the wards of an "average" asylum, to be shared by patients and married attendants (Bucknill, 1858).A cottage designed for this purpose had even attracted considerable public interest at the Paris Exhibition of 1867, and from 1893 a number of cottages shared by retired nurses with former in-patients were sponsored by the then Mental After-Care Association, of which Bucknill was one of the founders. His plans were met with both support and opposition and it is unclear why the disastrous policy of building more and larger asylums was adopted instead (McCandless, 1979).The potential of shared homes was re-discovered in the 1960s though without reference to the earlier ideas, and specialist housing played a key part in the establishment and progress ot