We investigated the effect of supplementing women weekly with 7000 microg retinol equivalents as preformed vitamin A or beta-carotene vs. a placebo, on the incidence of night blindness during pregnancy and the postpartum period in the rural plains of Nepal. Over a period of approximately 3 y, approximately 29,000 women of child-bearing age, living in 171 wards that were randomized to one of the three supplements, contributed 9932 first pregnancies. A prospective, weekly surveillance identified night blindness in pregnant women, verified further by detailed questioning about nighttime vision. After delivery, women were also interviewed at approximately 3 and approximately 6 mo postpartum to elicit a night blindness history over the preceding 3 mo. Vitamin A supplementation reduced the incidence of night blindness during pregnancy from 10.7% among controls to 6.7% (relative risk 0.62, 95% confidence interval: 0.45-0.85). beta-Carotene supplementation had less of an effect (0. 83, 0.63-1.11). Among women who took >95% of their vitamin A supplements during pregnancy, incidence of verified night blindness was reduced by 67%. Incidence (per 100 person-years) of night blindness during the first 3 mo postpartum was 11.3 in the control, 4.3 in the vitamin A and 8.7 in the beta-carotene groups, yielding corresponding relative risks of 0.38 (0.26-0.55) and 0.77 (0.57-1. 04). In the second 3 mo postpartum, both vitamin A and beta-carotene reduced night blindness by approximately 50%. Vitamin A intakes approaching a recommended amount for pregnancy markedly reduced but did not eliminate night blindness in Nepali women. Greater intakes of vitamin A than provided and/or other nutrients may be needed to prevent maternal night blindness in rural South Asia.