Humans who lack availability of carbohydrate fuels may provide important models for the study of physiological control mechanisms. We compared seven patients who had unavailability of muscle glycogen and blood glucose as oxidative fuels due to muscle phosphofructokinase deficiency (PFKD) with five patients who had a selective defect in long-chain fatty acid oxidation due to carnitine palmitoyltransferase deficiency (CPTD) and with six healthy subjects. Peak cycle exercise work rate, peak O2 uptake (Vo2), and arteriovenous O2 difference were markedly lower (P less than 0.001) for PFKD patients (23 +/- 6 W, 14 +/- 2 ml.min-1.kg-1, and 7.1 +/- 0.5 ml/dl, respectively) than for CPTD patients (142 +/- 33 W, 31 +/- 4 ml.min-1.kg-1, and 15.0 +/- 0.8 ml/dl, respectively) or healthy subjects (171 +/- 17 W, 36 +/- 1 ml.min-1.kg-1, and 16.4 +/- 0.7 ml/dl, respectively). Peak cardiac output (Q) was similar (P less than 0.05) in all three groups, but the slope of increase in Q (l/min) on Vo2 (l/min) from rest to exercise (delta Q/ delta Vo2) was more than twofold greater (P less than 0.001) for PFKD patients (11.2 +/- 1.2) than for CPTD patients (4.6 +/- 0.6) and healthy subjects (4.6 +/- 0.2). Increasing availability of blood-borne oxidative substrates capable of metabolically bypassing the defect at phosphofructokinase (by fasting plus prolonged moderate exercise to increase plasma free fatty acids or by iv lactate infusion) increased peak work rate, Vo2, and arteriovenous O2 difference, lacked consistent effect on peak Q, and normalized delta Q/ delta Vo2 in PFKD patients. The results extend our previous observations in patients with a block in muscle glycogen but not blood glucose oxidation due to phosphorylase deficiency and imply that specific unavailability of muscle glycogen as an oxidizable fuel is primarily responsible for abnormal muscle oxidative metabolism and associated exercise intolerance and exaggerated delta Q/ delta Vo2 in muscle PFKD. The findings also endorse the concept that factors closely linked with muscle oxidative phosphorylation participate in regulating delta Q/ delta Vo2, likely via activation of metabolically sensitive muscle afferents.