We determined changes in rat plantaris, diaphragm, and intercostal muscle metabolites following exercise of various intensities and durations, in normoxia and hypoxia (FIO2 = 0.12). Marked alveolar hyperventilation occurred during all exercise conditions, suggesting that respiratory muscle motor activity was high. [ATP] was maintained at rest levels in all muscles during all normoxic and hypoxic exercise bouts, but at the expense of creatine phosphate (CP) in plantaris muscle and diaphragm muscle following brief exercise at maximum O2 uptake (VO2max) in normoxia. In normoxic exercise plantaris [glycogen] fell as exercise exceeded 60% VO2max, and was reduced to less than 50% control during exhaustive endurance exercise (68% VO2max for 54 min and 84% for 38 min). Respiratory muscle [glycogen] was unchanged at VO2max as well as during either type of endurance exercise. Glucose 6-phosphate (G6P) rose consistently during heavy exercise in diaphragm but not in plantaris. With all types of exercise greater than 84% VO2max, lactate concentration ([LA]) in all three muscles rose to the same extent as arterial [LA], except at VO2max, where respiratory muscle [LA] rose to less than half that in arterial blood or plantaris. Exhaustive exercise in hypoxia caused marked hyperventilation and reduced arterial O2 content; glycogen fell in plantaris (20% of control) and in diaphragm (58%) and intercostals (44%). We conclude that respiratory muscle glycogen stores are spared during exhaustive exercise in the face of substantial glycogen utilization in plantaris, even under conditions of extreme hyperventilation and reduced O2 transport. This sparing effect is due primarily to G6P inhibition of glycogen phosphorylase in diaphragm muscle. The presence of elevated [LA] in the absence of glycogen utilization suggests that increased lactate uptake, rather than lactate production, occurred in the respiratory muscles during exhaustive exercise.