The responses to sublingual nifedipine (20 mg) and placebo were compared in normal subjects during two studies on cycle ergometer [progressive exercise and constant work-load exercise at approximately 60% of maximal O2 consumption (VO2max)]. The use of nifedipine did not modify maximal power, ventilation (VE), VO2, and heart rate (HR) at the end of the multistage progressive exercise (30-W increments every 3 min). Over the 45 min of the constant-load exercise and the ensuing 30-min recovery we observed with nifedipine compared with placebo 1) no differences in VO2, VE, respiratory exchange ratio, and systolic arterial blood pressure; 2) a higher HR (P less than 0.001) and lower diastolic arterial blood pressure (P less than 0.01); 3) a greater and more prolonged rise in norepinephrine (P less than 0.01) and growth hormone (P less than 0.001); 4) no significant differences in epinephrine and insulin and a lesser increase in glucagon during recovery (P less than 0.01); and 5) a lesser fall in blood glucose (P less than 0.01) and greater increase in acetoacetate (P less than 0.001), beta-hydroxybutyrate (P less than 0.05), and blood lactate (P less than 0.001). Our data do not support the hypothesis that nifedipine reduces hormonal secretions in vivo and are best explained by an enhanced secretion of catecholamines compensating for the primary vasodilator effect of nifedipine.