The effects of plasma drug binding on the relationship between propranolol concentration and the antagonism of isoproterenol tachycardia have been investigated in 8 normal subjects and 8 hypertensive patients. During chronic intravenous infusion of propranolol giving a narrow range of total plasma concentrations (22.5 to 50 ng/ml), there was, at best, a poor correlation with effects. On the other hand, there was excellent correlation between efficacy and free drug concentration, which fitted the predictions of the receptor theory of competitive antagonism. The true affinity constant for the binding of propranolol to its receptor can be calculated in terms offree drug concentration (KAfree) and was found to vary 2-fold compared to the a.ffinity constant in terms of total plasma concentration (KA4 total) which varied 4-fold, the greater variation being due to plasma binding differences. Compared to normal subjects, KAfree and KAtotal were smaller in hypertensive subjects, implying lesser sensitivity to the drug, and plasma propranolol binding was greater. There was no difference in KAfree between high- and low-renin essential hypertensives, but KAtotal was smaller in the high-renin group due to increased plasma binding which did not reach statistical significance. It is concluded that the effect of propranolol on heart rate is a predictable function offree drug concentration in man and that the contribution of individual variation in receptor sensitivity to differences in oral dosage requirement is minor compared to that of variations in bioavailability.