The pharmacokinetics of penbutolol 40 mg, its reduction in exercise-induced tachycardia, and the in vitro inhibition of radioligand binding to beta-adrenoceptors by plasma have been investigated in 7 healthy volunteers. The peak penbutolol concentration of 285 ng/ml was observed 1.2 h after administration, and the maximum of 4'-OH-penbutolol of 4.76 ng/ml was found after 1.64 h. Penbutolol was detected for up to 48 h, and 4'-OH-penbutolol dropped below the limit of detection after about 10 h. The terminal plasma concentration of penbutolol declined with an average half-life of 19 h. The maximum reduction in exercise-induced tachycardia was 33 beats/min 2.6 h after taking penbutolol. There was still a significant reduction of about 7 beats/min after 48 h. This effect could be adequately explained by the concentration-time course of penbutolol in combination with Clark's model of the concentration-effect relationship. Antagonist activity in plasma caused 91% inhibition of radioligand binding in vitro to beta 2-adrenoceptors on rat reticulocyte membranes 1.6 h after intake of penbutolol. By 48 h after intake, radioligand binding was still significantly inhibited (23%). The in vitro inhibition of radioligand binding by plasma showed a linear correlation with the reduction in exercise-induced tachycardia for all phases of the workload. The time course of the reduction in heart rate was completely explained by the in vitro inhibition of radioligand binding. However, it was not possible to explain the in vitro inhibition of radioligand binding by the concentration-time course of penbutolol using a simple competition model, although both variables were based on the same sampling site. When the in vitro inhibition of radioligand binding was plotted against the penbutolol concentration at the same sampling times (with both variables transformed to multiples of the apparent inhibition constant) the discrepancy became even more apparent as time-related counterclockwise hysteresis. None of the known metabolites of penbutolol can explain the discrepancy between the penbutolol concentration and the inhibition of radioligand binding in vitro. It appears that an other active metabolite is formed, which contributes to the effect in vitro and in vivo and so can explain the observed discrepancy.