ABSTRACT: BACKGROUND:Few studies have identified modifiable risk factors that are associated with the prevention of preterm delivery (PTD). This study examined the relationship between PTD and physical activity during pregnancy. METHODS:Data were obtained by medical record review and postpartum questionnaires from a cohort of African American women (N = 1,410) delivering singleton infants. Physical activity was self-reported and analyses compared any and none. Additional analyses classified leisure time physical activity (LTPA) and walking for a purpose as 0, 1 to 19, 20 to 39, and 40 minutes per day or more and stair climbing as 0, 1 to 5, 6 to 9, and 10 or more times per day. Log-Poisson models adjusted for previous PTD, pregnancy complications, and income were used to examine the association between PTD and physical activity during pregnancy across body mass index categories. RESULTS:Overall, 16.4% of deliveries were preterm. LTPA was associated with a decreased prevalence of PTD (prevalence ratio [PR], 0.73; 95% confidence interval [CI], 0.55-0.96), but stratification by maternal prepregnancy body mass index suggested that LTPA was only protective against PTD among women with normal weight (PR, 0.43; 95% CI, 0.23-0.79). Stair climbing 10 or more times per day was associated with a decreased prevalence of PTD among women with normal weight (PR, 0.32; 95% CI, 0.11-0.94) and women with overweight (PR, 0.24; 95% CI, 0.07-0.80) only. Walking for a purpose (e.g., to the store, the bus stop, or to work) was not associated with PTD. CONCLUSIONS:African American women who participate in either LTPA or stair climbing during pregnancy have a decreased prevalence of PTD, but the protective effect varied by maternal body mass index.