Context: Prior study suggests that contextual characteristics of medical schools (e.g., state demographics, public vs private, NIH research effort) predict output of rural physicians without also considering the effects of the medical schools' own policies and programs. Purpose: This study examines medical school commitment to rural policies and programs and its relationship to contextual characteristics and rural physician output. Methods: A survey of 122 US allopathic medical schools provided data to construct a 32-item Rural Commitment Index for each medical school. Data for other characteristics were linked from published sources. Correlations, t tests, and multiple regression analysis were used to study the association between variables and percentage of medical school graduates (1988-1996) who were in rural primary care practice in 2000. Findings: Among 90 medical schools (response rate, 73.8%), the Rural Commitment Index correlated with the percentage of the state population that is rural and whether the school is public or private, and it joined percentage state population rural, public vs private, and National Institutes of Health support in correlating with percentage of graduates in rural primary care. In a regression model that explained 48.4% of variation in the percentage of graduates in rural primary care, the Rural Commitment Index explained most variation, followed by percentage state population rural, public vs private, National Institutes of Health support, and the interaction between the Rural Commitment Index and public vs private. Conclusions: The findings support the proposition that observable institutional commitment affects rural physician output and provide justification for a definitive study to verify that a change in medical school commitment to rural medicine produces a change in rural physician output.