BACKGROUND: There is a high incidence of impairments among working age adults. Many countries offer personal assistance in the form of individualised support for people living in the community by a paid assistant other than a healthcare professional for at least 20 hours per week. OBJECTIVES: To assess the effectiveness of personal assistance for adults with physical impairments, and the impacts of personal assistance on others, compared to other interventions. SEARCH STRATEGY: Electronic databases including CENTRAL, MEDLINE, EMBASE, CINAHL, PsycINFO, ERIC, Dissertation Abstracts International and a variety of specialist Swedish databases were searched from 1980 to June 2005; reference lists were checked; 345 experts, organisations, government bodies and charities were contacted in an attempt to locate relevant research. SELECTION CRITERIA: Adults (19-64) with physical impairments living in the community who require assistance to perform tasks of daily living (e.g., bathing and eating) and participate in normal activities due to permanent impairments. Controlled studies of personal assistance in which participants were prospectively assigned to study groups and in which control group outcomes were measured concurrently with intervention group outcomes were included. DATA COLLECTION AND ANALYSIS: Titles and abstracts were examined by two reviewers. Outcome data were extracted. Studies were assessed for the possibility of bias. Results and potential sources of bias are presented for included studies. MAIN RESULTS: One randomised controlled trial involving 817 participants compared personal assistance versus usual care was identified. Whilst personal assistance was generally preferred over other services, some people prefer other models of care. This review indicates that personal assistance may have some benefits for some recipients and may benefit caregivers. Whilst paid assistance probably substitutes for informal care and may cost government more than alternatives, the total costs to recipients and society are currently unknown. AUTHORS' CONCLUSIONS: Research in this field is limited. When implementing new programmes, recipients could be randomly assigned to different forms of assistance. While advocates may support personal assistance for myriad reasons, this review demonstrates that further studies are required to determine which models of personal assistance are most effective and efficient for particular people.