This study examines the distances traveled for inpatient obstetrics care by women residing in rural Alabama in 1983 and 1988. During that time 23 rural hospitals in the state stppped providing obstetrics services and mean travel distances increased by 6.8 miles. However, in 1988 50% of rural pregnant women bypassed the nearest rural hospital still providing obstetrics services. Multivariate techniques are used to examine the effects of distance and service offerings of rural hospitals and their substitutes on the actual distance traveled for care. Patient characteristics are also considered. The most important finding is that a 5% increase in per capita income in the woman’s home county is associated with a 20% increase in actual travel distance, other things equal. Implications for rural health policy are discussed. © Lippincott-Raven Publishers.