Objectives: The purpose of this study was to determine the effect of regular leisure physical activity (RLPA) on two different adverse birth outcomes: timeliness of delivery (<37 weeks, preterm; 37–42 weeks, term; and >42 weeks, postterm) and low birth weight (<1500 g, very low; 1500–2499 g, low). Methods: The present sample, consisting of 9089 women, was obtained from the 1988 National Maternal and Infant Health Survey (NMIHS) data. The NMIHS was developed to examine adverse birth outcomes by assessing various maternal characteristics such as demographic, behavioral, and health care factors not found in vital statistics data. Results: The specified adjusted models obtained by logistic regression indicate that women who failed to engage in RLPA before and during their pregnancy were more likely to give birth to a very low birth weight baby [OR = 1.75; 95% CI (1.50, 2.04)] but not to a low birth weight baby [OR = 1.15; 95% CI (0.99, 1.34)] compared with women who remained active before and during pregnancy. Moreover, previously active women who stopped physical activity during pregnancy were more likely to give birth to a low birth weight [OR = 1.28; 95% CI (1.05, 1.56)] or a very low birth weight [OR = 2.05; 95% CI (1.69, 2.48)] baby than women who remained active before and during pregnancy. There was no significant relationship between RLPA and timeliness of delivery. Similar results were found after controlling for maternal medical risk. Conclusions: RLPA during pregnancy had no deleterious effect on birth outcomes among these women. RLPA may also reduce the risk of low birth weight outcomes.