Objective: We sought to summarise the evidence for an association between rural background and rural practice by systematically reviewing the national and international published reports. Design: A systematic review. Setting: A search of the national and international published reports from 1973 to October 2001. Subject: The search criteria included observational studies of a case-control or cohort design making a clear and quantitative comparison between current rural and urban doctors, this resulted in the identification of 141 studies for potential inclusion. Results: We systematically reviewed 12 studies. Rural background was associated with rural practice in 10 of the 12 studies, in which it was reported, with most odds ratios (OR) approximately 2-2.5. Rural schooling was associated with rural practice in all 5 studies that reported on it, with most OR approximately 2.0. Having a rural partner was associated with rural practice in 3 of the 4 studies reporting on it, with OR approximately 3.0. Rural undergraduate training was associated with rural practice in 4 of 5 studies, with most OR approximately 2.0. Rural postgraduate training was associated with rural practice in 1 of 2 studies, with rural doctors reporting rural training about 2.5 times more often. Conclusions: There is consistent evidence that the likelihood of working in rural practice is approximately twice greater among doctors with a rural background. There is a smaller body of evidence in support of the other rural factors studied, and the strength of association is similar to that for rural background.