Objective: To review studies reporting on the effectiveness of psychiatry service delivery for older people and people with dementia in hospital and residential aged care. Methods: A systematic search of four databases was conducted to obtain peer-reviewed literature reporting original research published since June 2004 evaluating a psychiatry service for older people (aged 60 years and over) or people with dementia in inpatient or residential aged care settings. Results: From the 38 included studies, there was consistent low-to-moderate quality evidence supporting the effectiveness of inpatient older persons' mental health wards (n = 14) on neuropsychiatric symptoms, mood, anxiety and quality of life. Inpatient consultation/liaison old age psychiatry services (n = 9) were not associated with improved depression, quality of life or mortality in high-quality randomised studies. However, low-quality evidence demonstrated improved patient satisfaction with care and reduced carer stress. The highest quality studies demonstrated no effect of psychiatric in-reach services to residential aged care (n = 9) on neuropsychiatric symptoms but a significant reduction in depressive symptoms among people with dementia. There was low-quality evidence that long-stay intermediate care wards (n = 6) were associated with reduced risk for dangerous behavioural incidents and reduced costs compared to residential aged care facilities. There was no effect of these units on neuropsychiatric symptoms or carer stress. Conclusions and implications: The scarcity of high-quality studies examining the effectiveness of old age psychiatry services leaves providers and policy-makers to rely on low-quality evidence when designing services. Future research should consider carefully which outcomes to include, given that staff skill and confidence, length of stay, recommendation uptake, patient- and family-reported experiences, and negative outcomes (i.e. injuries, property damage) are as important as clinical outcomes.