Glucose infusion can prevent the increase in glucose production (Ra) and increase glucose uptake (Rd) during exercise of moderate intensity. We postulated that 1) because in postabsorptive intense exercise (>80% maximal O2 uptake) the eightfold increase in Ra may be mediated by catecholamines rather than by glucagon and insulin, exogenous glucose infusion would not prevent the Ra increment, and 2) such infusion would cause greater Rd. Fit young men were exercised at >85% maximal O2 uptake for 14 min in the postabsorptive state [controls (Con), n = 12] or at minute 210 of a 285-min glucose infusion. In seven subjects, the infusion was constant (CI; 4 mg . kg-1 . min-1), and in seven subjects it was varied (VI) to mimic the exercise Ra response in Con. Although glucose suppressed Ra to zero (with glycemia approximately 6 mM and insulin approximately 150 pM), an endogenous Ra response to exercise occurred, to peak increments two-thirds those in Con, in both CI and VI. Glucagon was unchanged, and very small increases in the glucagon-to-insulin ratio occurred in all three groups. Catecholamine responses were similar in all three groups, and correlation coefficients of Ra with plasma norepinephrine and epinephrine were significant in all. In all CI and VI, Rd at rest was 2x Con, increased earlier in exercise, and was higher for the 1 h of recovery with glucose infusion. Thus the Ra response was only partly attenuated, and the catecholamines are likely to be the regulators. This suggests that an acute endogenous Ra rise is possible even in the postprandial state. Furthermore, the fact that more circulating glucose is used by muscle during exercise and early recovery suggests that muscle glycogen is spared.