To test the hypothesis that increments in plasma cyclic AMP during beta-adrenergic stimulation reflect integrated second messenger function of the tissues activated by the agonist, graded adrenaline infusion resulting in plasma adrenaline concentrations within the physiological range was performed in 8 healthy subjects with and without concomitant beta-adrenoceptor blockade by iv propranolol. A significant correlation was found between increments in plasma adrenaline and plasma cyclic AMP in the experiments without beta-blockade; during concomitant beta-blockade the increase in plasma cyclic AMP concentrations at low adrenaline infusion rates was prevented, whereas a small increase in cyclic AMP was found at high adrenaline infusion rates, probably owing to incomplete beta-receptor blockade. Likewise, the adrenaline-induced increments in blood substrates (glucose, lactate, glycerol and beta hydroxybutyric acid) were significantly reduced but not completely prevented by beta-blockade. We conclude that an altered relationship between beta-agonist concentrations and plasma cyclic AMP may provide evidence for the existence of differences in beta-adrenergic sensitivity in man.