The mechanisms of postprandial glucose counterregulation-those that blunt late decrements in plasma glucose, prevent hypoglycemia, and restore euglycemia-have not been fully defined. To begin to clarify these mechanisms, we measured neuroendocrine and metabolic responses to the ingestion of glucose (75 g), xylose (62.5 g), mannitol (20 g), and water in ten normal human subjects to determine for each response the magnitude, temporal relationships, and specificity for glucose ingestion. Measurements were made at 10-min intervals over 5 h. By multivariate analysis of variance, the plasma glucose (P < 0.0001), insulin (P < 0.0001), glucagon (P < 0.03), epinephrine (P < 0.0004), and growth hormone (P < 0.01) curves, as well as the blood lactate (P < 0.0001), glycerol (P < 0.001), and beta-hydroxybutyrate (P < 0.0001) curves following glucose ingestion differed significantly from those following water ingestion. However, the growth hormone curves did not differ after correction for differences at base line. In contrast, the plasma norepinephrine (P < 0.31) and cortisol (P < 0.24) curves were similar after ingestion of all four test solutions, although early and sustained increments in norepinephrine occurred after all four test solutions. Thus, among the potentially important glucose regulatory factors, only transient increments in insulin, transient decrements in glucagon, and late increments in epinephrine are specific for glucose ingestion. They do not follow ingestion of water, xylose, or mannitol. Following glucose ingestion, plasma glucose rose to peak levels of 156+/-6 mg/dl at 46+/-4 min, returned to base line at 177+/-4 min, reached nadirs of 63+/-3 mg/dl at 232+/-12 min, and rose to levels comparable to base line at 305 min, which was the final sampling point. Plasma insulin rose to peak levels of 150+/-17 muU/ml (P < 0.001) at 67+/-8 min. At the time glucose returned to base line, insulin levels (49+/-12 muU/ml) remained fourfold higher than base line (P < 0.01); thereafter they declined but never fell below base line. Plasma glucagon decreased from 95+/-14 pg/ml to nadirs of 67+/-11 pg/ml (P < 0.001) at 84+/-9 min and then rose progressively to peak levels of 114+/-17 pg/ml (P < 0.001 vs. nadirs) at 265+/-12 min. Plasma epinephrine, which was 18+/-4 pg/ml at base line, did not change initially and then rose to peak levels of 119+/-20 pg/ml (P < 0.001) at 271+/-13 min. These data indicate that the glucose counterregulatory process late after glucose ingestion is not solely due to the dissipation of insulin and that sympathetic neural norepinephrine, growth hormone, and cortisol do not play critical roles. They are consistent with, but do not establish, physiologic roles for the counterregulatory hormones-glucagon, epinephrine, or both-in that process.