ABSTRACT: BACKGROUND:Low birth weight (LBW) is an important public health problem in undernourished populations. OBJECTIVE:We tested whether improving women's dietary micronutrient quality before conception and throughout pregnancy increases birth weight in a high-risk Indian population. DESIGN:The study was a nonblinded, individually randomized controlled trial. The intervention was a daily snack made from green leafy vegetables, fruit, and milk (treatment group) or low-micronutrient vegetables (potato and onion) (control group) from ? 90 d before pregnancy until delivery in addition to the usual diet. Treatment snacks contained 0.69 MJ of energy (controls: 0.37 MJ) and 10-23% of WHO Reference Nutrient Intakes of ?-carotene, riboflavin, folate, vitamin B-12, calcium, and iron (controls: 0-7%). The primary outcome was birth weight. RESULTS:Of 6513 women randomly assigned, 2291 women became pregnant, 1962 women delivered live singleton newborns, and 1360 newborns were measured. In an intention-to-treat analysis, there was no overall increase in birth weight in the treatment group (+26 g; 95% CI: -15, 68 g; P = 0.22). There was an interaction (P < 0.001) between the allocation group and maternal prepregnant body mass index (BMI; in kg/m(2)) [birth-weight effect: -23, +34, and +96 g in lowest (<18.6), middle (18.6-21.8), and highest (>21.8) thirds of BMI, respectively]. In 1094 newborns whose mothers started supplementation ? 90 d before pregnancy (per-protocol analysis), birth weight was higher in the treatment group (+48 g; 95% CI: 1, 96 g; P = 0.046). Again, the effect increased with maternal BMI (-8, +79, and +113 g; P-interaction = 0.001). There were similar results for LBW (intention-to-treat OR: 0.83; 95% CI: 0.66, 1.05; P = 0.10; per-protocol OR = 0.76; 95% CI: 0.59, 0.98; P = 0.03) but no effect on gestational age in either analysis. CONCLUSIONS:A daily snack providing additional green leafy vegetables, fruit, and milk before conception and throughout pregnancy had no overall effect on birth weight. Per-protocol and subgroup analyses indicated a possible increase in birth weight if the mother was supplemented ? 3 mo before conception and was not underweight. This trial was registered at www.controlled-trials.com/isrctn/ as ISRCTN62811278.