Background Childbirth is the most common reason for hospitalization in the United States. Assessing obstetric care quality is critically important for patients, clinicians, and hospitals in rural areas. Methods The study used hospital discharge data from the Statewide Inpatient Databases, Healthcare Cost and Utilization Project, Agency for Healthcare Research and Quality, for 9 states (Colorado, Iowa, Kentucky, New York, North Carolina, Oregon, Vermont, Washington, and Wisconsin) to identify all births in rural hospitals with 10 or more births/year in 2002 (N = 94,356) and 2010 (N = 103,880). Multivariate logistic regression was used to assess the relationship between hospital annual birth volume, measured as low (10‐110), medium (111‐240), medium‐high (241‐460) or high (>460), and 3 measures of obstetric care quality (low‐risk cesarean rates for term, vertex, and singleton pregnancies with no prior cesarean; nonindicated cesarean; and nonindicated induction) and 2 patient safety measures (episiotomy and perineal laceration). Results The odds of low‐risk and nonindicated cesarean were lower in medium‐high and high‐volume rural hospitals compared with low‐volume hospitals after controlling for maternal demographic and clinical factors. In low‐volume hospitals, odds of labor induction without medical indication were higher than in medium‐volume hospitals, but not significantly different from medium‐high or high‐volume hospitals. Odds of episiotomy were greater in medium‐high or high‐volume hospitals than in low‐volume hospitals. The likelihood of perineal laceration did not differ significantly by birth volume. Conclusions Obstetric quality and safety outcomes vary significantly across rural hospitals by birth volume. Better performance is not consistently associated with either lower or higher volume facilities.