Background During 2013-2016 the specialist continuing care service at South London and Maudsley NHS Foundation Trust (SLaM) underwent significant changes. Patients receiving care from this service are an under researched group, and there is a lack of evidence available for different models of service delivery. Previous research on the discharge of elderly long stay psychiatric patients was mainly conducted during the deinstitutionalisation movement and reported that this group generally has good outcomes following discharge. This study aimed to explore the effects on quality and costs of care of a changing service model, as well as the effects of discharge from longterm inpatient care of a vulnerable group of older people during a planned period of bed reductions. Method: This prospective longitudinal evaluation measured quality and costs of care for patients within the specialist care units at SLaM during a period of programmed discharges and specialist care unit (SCU) closures. The primary outcome measure used was the NPI-NH. Secondary outcome measures were the BADLS and the QUALIDEM. Costs were measured using the CSRI. Additional data were collected on safety incidents, mortality, and staff sickness, as well as demographic data on study participants. Results: The population remaining in the SCUs had significantly higher levels of behavioural symptoms and dependency, and lower quality of life at the end of the study. The number of safety incidents occurring at the SCUs decreased significantly. Although there were no statistically significant differences in mortality rates, a high proportion of discharged participants died prior to six-month follow-up. Participants had higher dependency levels following discharge. Costs of care were significantly less for participants that were discharged into community placements. Although costs of care at the patient level increased in the SCUs by the end of the study, overall costs decreased. Conclusions: The specialist continuing care service now cares for a more complex group of individuals, which has had impacts on the costs of running the service. The service is safer as measured by reported risk incidents. Some negative impacts on those discharged from the service are apparent although the study is limited by its small sample size. Care was significantly cheaper for participants following discharge to community settings. Caution should be taken when discharging long stay patients from the SCU setting and further research is needed on longer term outcomes for this group, as well as an evaluation of the community services that have replaced the SCUs.