Normal aging is characterized by a progressive impairment in glucose tolerance. An important mechanism underlying the glucose intolerance of aging is an impairment in glucose-induced insulin release. These studies were conducted to determine whether the age-related impairment in insulin release was caused by a decreased beta-cell sensitivity to glucose-dependent insulinotropic polypeptide (GIP). Thirty-one Caucasian men were divided into four groups: two young groups (age range: 19-26 yr, n = 15) and two old groups (age range: 67-79 yr, n = 16). Each volunteer participated in three studies (n = 93 clamps). Hyperglycemic clamps were conducted at two doses [basal plasma glucose (G) + 5.4 mmol/L and G + 12.8 mmol/L] for 120 min. In the initial hyperglycemic clamp, only glucose was infused. In subsequent studies, GIP was infused at a final rate of 2 or 4 pmol/ kg(-1) x min(-1) from 60-120 min. Basal plasma insulin and GIP levels were similar in the young (41 +/- 6 and 51 +/- 6 pmol/L) and the old subjects (42 +/- 6 and 66 +/- 12 pmol/L) in all studies. First- and second-phase insulin responses were similar during the control study and during the first 60 min of each GIP infusion study in both groups. The 90-120 min GIP values were similar between groups and between hyperglycemic plateaus during the 2 and 4 pmol/kg(-1) x min(-1) GIP infusion (young: 342 +/- 28 and 601 +/- 44 pmol/L, old: 387 +/- 45 and 568 +/- 49 pmol/L). In response to the GIP infusions, significant increases in insulin occurred in young and old at both glucose levels (P < 0.01). The potentiation of the insulin response caused by GIP was greater in the young subjects than in the old, in the G + 5.4 mmol/L studies (P < 0.05). However, the insulin response to GIP was similar in both young and old during the G + 12.8 mmol/L clamps. The insulinotropic effect of the incretin was higher in the young and in the old, in the G + 12.8 mmol clamps, than in the G + 5.4 mmol/L clamps. We conclude that normal aging is characterized by a decreased beta-cell sensitivity to GIP during modest hyperglycemia, which may explain, in part, the age-related impairment in glucose-induced insulin release. We also find that the insulinotropic effect of GIP is increased with increasing levels of hyperglycemia.