This study examined the physiological effect of reduced plasma cortisol (C) during prolonged exercise in humans. The effects of normal C (NC) were compared with metyrapone-induced low C (LC) on plasma substrate availability and the respiratory exchange ratio during 2 h of exercise at approximately 60% peak O2 consumption in nine subjects. The C responses were compared with preexercise (Pre) levels and with a rest day (Con). At rest, C was attenuated by approximately 70% for LC compared with NC. At rest, plasma glucose, lactate, glycerol, beta-hydroxybutyrate, alanine, branched-chain amino acids, insulin, glucagon, growth hormone, epinephrine, and norepinephrine were similar under LC and NC (P > 0.05). During exercise under NC, plasma C increased compared with Pre, whereas it remained unchanged during LC. During NC, plasma C was elevated at 90 min (compared with Con) and at 120 min (compared with Con and Pre). During exercise, plasma glucose decreased to the same extent and lactate was similar under both conditions, whereas plasma glycerol, beta-hydroxybutyrate, alanine, and branched-chain amino acids were higher (P < 0.01) under NC. Plasma insulin declined (P = 0.01) to a greater extent under LC, whereas growth hormone, epinephrine, and norepinephrine tended to be higher (0.05 </= P </= 0.10). Plasma glucagon increased under both conditions (P < 0.01). The respiratory exchange ratio did not differ between conditions. We conclude that, during exercise, 1) C accelerates lipolysis, ketogenesis, and proteolysis; 2) under LC, glucoregulatory hormone adjustments maintain glucose homeostasis; and 3) LC does not alter whole body substrate utilization or the ability to complete 2 h of moderate exercise.