Background: Although perinatal walking and yoga have been associated with decreased risks of pregnancy complications, associations with offspring birth size have been inconsistent. We investigated associations of prepregnancy and early pregnancy leisure-time light/moderate walking and yoga practice with birth size. Methods: Study participants (N = 3687) reported leisure-time physical activity duration (hours per week) in the year before pregnancy and early pregnancy. Birth size was abstracted from medical records. Regression was used to determine mean differences in birth weight, head circumference, and ponderal index. Interaction terms were used to assess effect modification by offspring sex. Results: About one-third of women reported light/moderate leisure-time walking and about 10% reported yoga practice. Women in the highest tertile for prepregnancy (mean: 2.9 h/wk; range: 1.4–20 h/wk) or early pregnancy (mean: 5.9 h/wk; range: 3.1–24 h/wk) light/moderate walking had offspring with 0.9 and 1.5 kg/m3 greater ponderal index (95% confidence interval, 0.3 to 1.4 and 0.7 to 2.4, respectively) compared with women who reported no light/moderate walking in the same time period. Light/moderate walking was not associated with birth weight or head circumference. Yoga practice was not associated with birth size. Associations were similar by offspring sex. Conclusion: Light/moderate leisure-time walking may be associated with greater offspring ponderal index.