This study investigated the hypothesis that training-induced reductions in exercise blood glucose utilization can occur independently of increases in muscle mitochondrial potential. To induce a training adaptation, eight active participants (23 +/- 1 yr, 80.6 +/- 3.7 kg, mean +/- SE) with a maximal oxygen consumption (VO2max) of 45.5 +/- 2.4 ml.kg-1.min-1, cycled at 59% VO2max for 2 h per day for 10 consecutive days. Measurements of blood glucose appearance (Ra) and disappearance (Rd), using a primed continuous infusion of [6,6-2H2]glucose, were made during 90 min of cycle exercise (59% VO2max) performance before and after training. Training resulted in a 25% decrease (P < 0.01) in mean glucose Ra during exercise (43.0 +/- 3.7 to 34.4 +/- 2.8 mumol.kg-1.min-1). Since blood glucose concentration was not different between training conditions, glucose metabolic clearance rate was also depressed (P < 0.05). Exercise-induced glycogen depletion in vastus lateralis muscle was reduced (P < 0.05) with training. Calculation of carbohydrate and fat oxidation based on the respiratory exchange ratio supported a shift toward greater preference for fat. Because training did not elicit changes in the maximal activities of citrate synthase and malate dehydrogenase, two enzymes of the citric acid cycle, it would appear that increases in mitochondrial potential are not necessary for the adaptations that occurred.