Studies conducted over the last 25 years have suggested that a substantial fraction of hospital use is inappropriate. This report examines the relationship between patient and provider (e.g., physician and hospital) characteristics and inappropriate hospital use, in terms of both admissions and days. The study examines adult nonpregnancy hospitalizations from a randomized trial of health insurance, the RAND Health Insurance Experiment. The data include evaluations of the appropriateness of inpatient treatment based on medical record review, patient characteristics (sociodemographic, economic, and health status), and provider characteristics (specialty and practice patterns for physicians, accreditations and facilities for hospitals). The findings indicate that women and patients of older physicians are more likely to have been inappropriately admitted to the hospitals. The authors conclude that patient and provider characteristics influence the likelihood that an admission is inappropriate; however, differences in provider and patient characteristics do not account for geographic differences in inappropriate hospitalization.