ABSTRACT: BACKGROUND:Recent developments within the United Kingdom's (UK) health care system have re-awakened interest in community hospitals (CHs) and their role in the provision of health care. This integrative literature review sought to identify and assess the current evidence base for CHs. METHODS:A range of electronic reference databases were searched from January 1984 to either December 2004 or February 2005: Medline, Embase, Web of Knowledge, BNI, CINAHL, HMIC, ASSIA, PsychInfo, SIGLE, Dissertation Abstracts, Cochrane Library, Kings Fund website, using both keywords and text words. Thematic analysis identified recurrent themes across the literature; narrative analyses were written for each theme, identifying unifying concepts and discrepant issues. RESULTS:The search strategy identified over 16,000 international references. We included papers of any study design focussing on hospitals in which care was led principally by general practitioners or nurses. Papers from developing countries were excluded. A review of titles revealed 641 potentially relevant references; abstract appraisal identified 161 references for review. During data extraction, a further 48 papers were excluded, leaving 113 papers in the final review. The most common methodological approaches were cross-sectional/descriptive studies, commentaries and expert opinion. There were few experimental studies, systematic reviews, economic studies or studies that reported on longer-term outcomes. The key themes identified were origin and location of CHs; their place in the continuum of care; services provided; effectiveness, efficiency and equity of CHs; and views of patients and staff. In general, there was a lack of robust evidence for the role of CHs, which is partly due to the ad hoc nature of their development and lack of clear strategic vision for their future. Evidence for the effectiveness and efficiency of the services provided was limited. Most people admitted to CHs appeared to be older, suggesting that admittance to CHs was age-related rather than condition-related. CONCLUSION:Overall the literature surveyed was long on opinion and short of robust studies on CHs. While lack of evidence on CHs does not imply lack of effect, there is an urgent need to develop a research agenda that addresses the key issues of health care delivery in the CH setting.