The authors studied whether the fatiguing effects of eating lunch are greater for carbohydrate-rich meals than for other meals, and related the time course of behavioral change to plasma glucose, insulin, and amino acids. On different occasions, in counterbalanced order, normal women (N = 7) fasted overnight, ate a standard breakfast, and at lunch either continued to fast or ate a high-carbohydrate, low-protein meal; a hedonically similar meal containing both carbohydrate and protein; or a high-protein, low-carbohydrate meal. Meals were isocaloric and equated for fat content. Only the carbohydrate meal significantly increased fatigue, which could not be attributed to hypoglycemia because plasma glucose remained elevated. Fatigue began approximately, when the carbohydrate meal elevated the plasma tryptophan ratio but ended even though the ratio remained elevated. Fatigue after a high-carbohydrate lunch could not be explained by reactive hypoglycemia or sweet taste, and could partially be explained by the hypothesis that fatigue parallels an elevation of the tryptophan ratio.