Preterm delivery rates have remained consistently higher in the US than France, but the reasons for this excess remain poorly understood. We examined if differences in socio-demographic risk factors or more liberal use of obstetrical interventions contributed to higher rates in the US. Data on singleton live births in 1995, 1998 and 2003 from US birth certificates and the French National Perinatal Survey were used to analyze preterm delivery rate by maternal characteristics (age, parity, marital status, education, race (US)/nationality (France), prenatal care and smoking). We distinguished between preterm deliveries with a cesarean or a labor induction and those without these interventions. Unadjusted and adjusted risk ratios (RR) for the US compared to France were estimated using log-binomial regression. Preterm delivery rates were 7.9 % in the US and 4.7 % in France (risk ratio [RR] = 1.7, 95 % confidence interval [CI] 1.6–1.8). The US had more teen mothers and late entry to prenatal care, but fewer women smoked, although adjustment for these and other confounders did not reduce RR (1.8, 95 % CI 1.7–1.9). Preterm delivery rates associated with labor induction or cesarean were 3.3 % in the US and 2.1 % in France (RR 1.6, 95 % CI 1.5–1.7); the corresponding rates for preterm delivery without these interventions were 4.5 and 2.5 % (RR 1.8, 95 % CI 1.7–1.9), respectively. Key socio-demographic risk factors and more obstetric intervention do not explain higher US preterm delivery rates. Avenues for future research include the impact of universal access to health services (universal health insurance?) on health care quality and the association between more generous social policies, stress and the risks of preterm delivery.