The issue of rural hospital closings in the United States in recent years has become of increasing concern to health care policy analysts. Rural communities face unique health needs, necessitating access to local health care. Much has been written about the social, economic, legislative, and technological changes that have increased the stress on rural hospitals in the 1980s. However, quantifiable models have been lacking with which to examine in detail factors associated with rural hospitals and to correlate such factors with individual hospitals' risks of closing. In this study, we identify variables correlated with rural community hospital closures in the period 1980-87. Using epidemiologic case-control methods, 161 closed rural hospitals were matched 1 to 3 with a control group of 483 rural hospitals which remained open during the same period. A series of hospital performance indicators and demographic, economic, and social community variables were entered into a multiple logistic regression model. Four variables were found to be positively correlated with risk of closure. They are for-profit ownership; nongovernment, not-for-profit ownership; presence of a skilled nursing or other longterm care unit; and the number of other hospitals in the county. Variables negatively correlated with risk of closure were accreditation by the Joint Commission on the Accreditation of Healthcare Organizations, the number of facilities and services, and membership in a multihospital system. Policy and research implications at the Federal, State, and local levels are discussed.