Two hundred three (40 or 41/treatment, average parity 4.3) Large White x Landrace or Large White x Chester White x Landrace sows were used to determine the valine requirement of sows with a genetic capacity for high milk production. All diets were formulated to contain .90% total lysine, with all amino acids other than valine formulated to be at least 110% of their suggested estimates relative to lysine based on ratios derived from the National and Agricultural Research Councils. The control diet was formulated to .75% total valine, and crystalline valine replaced cornstarch to provide additional treatments containing .85, .95, 1.05, and 1.15% total dietary valine. Corresponding valine:lysine ratios were 83, 94, 106, 117, and 128%. Mean litter size after adjustment was 10.3 pigs across treatments, and average lactation length was 26 d. Number of pigs weaned was not affected by dietary valine (x̄ = 10.2 pigs), nor was daily sow feed (x̄ = 6.24 kg) or lysine (x̄ = 56 g) intake. Valine intake increased (linear, P < .001) as dietary valine increased. Litter weight at d 21 and at weaning (d 26) increased (linear, P < .02) with increasing dietary valine (62.4 to 65.5 kg and 76.1 to 79.9 kg, respectively). Litter weight gain increased from d 0 to 7 (linear, P < .06) and from d 0 to 21 and d 0 to weaning (linear, P < .02) as dietary valine increased. Dietary valine had no effect (P > .10) on sow weight change, 10th rib, or last lumbar backfat change from d 0 to 21 or d 0 to weaning or on days from weaning to estrus. These results demonstrate that increasing dietary valine for high-producing sows (21-d litter weights > 60 kg) results in improved litter weight gain. Based on the linear responses observed, the requirement is at least 1.15% of the diet (72 g/d of valine intake) to maximize litter weaning weight and litter weight gain, much greater than recommended currently by the National Research Council (100% of lysine, 36.5 g/d) or the Agricultural Research Council (70% of lysine, 25.5 g/d).