Background: The current rate of cesarean delivery in the United States is 31 percent. Previous studies have suggested that exercise during pregnancy may be associated with a lower risk of cesarean delivery, but sample sizes were small and methods often inadequate. This study examined whether or not an association exists between prenatal exercise and delivery mode using data from the 2004 and 2005 North Carolina Pregnancy Risk Assessment Monitoring System (PRAMS) survey. Methods: PRAMS postpartum questionnaire responses about frequency of exercise during the last 3 months of pregnancy for 1,955 women without a prior cesarean delivery were linked to birth certificates. Results: Among 1,342 women delivering at term, exercise was not associated with delivery mode in this data set: compared with women exercising less than once a week, neither women exercising one to four times per week nor those exercising five times or more per week had an altered risk of cesarean (risk ratio [RR] [95% confidence limit] [CL] 0.89 [0.69–1.15], 1.04 [0.66–1.64], respectively, adjusted for parity, gestational age, hypertension). Among 613 women delivering preterm, the results were also not statistically significant, but a compelling trend toward a protective effect could be seen (RR [95% CL] 0.65 [0.38–1.13], 0.62 [0.29–1.33]). Conclusions: Maternal self-reported frequency of exercise during pregnancy was not associated with a reduced risk of cesarean delivery. Larger studies with better exposure ascertainment may provide a more definitive answer.