Résumé : Summary: Experiments were performed on canine and human isolated coronary arteries to characterise human coronary beta adrenoceptors and to determine whether or not beta blocking agents with different ancillary properties unmask the alpha adrenergic effect of noradrenaline in a similar way. The inhibitory effects of atenolol (a beta1 selective antagonist), epanolol (a beta1 selective antagonist with modest intrinsic sympathetic activity), and propranolol (a non-selective antagonist) were assessed on isoproterenol concentration-response curves. Regression analysis provided slopes not significantly different from unity and similar pA2 values for each agent in both preparations. In a second group of experiments, the effects of noradrenaline (10 μmol·litre-1) were assessed in the absence and presence of beta blockade. At equipotent doses (1 log or 2 log units from the pA2 values) each beta blocking agent unmasked the alpha effect of noradrenaline in the same way. This alpha effect of noradrenaline (10 μmol·litre-1) was completely abolished by prazosin 1 μmol·litre-1 in canine coronary arteries but only partially antagonised in human coronary arteries.Thus the property of a beta blocking agent to unmask the alpha adrenergic effect of adrenaline is mainly related to its affinity for the coronary smooth muscle beta adrenoceptors. These beta adrenoceptors were very similar in both preparations and appear to be mainly beta1. The alpha adrenoceptors seem, nevertheless, to be different and resistant to prazosin in human preparations.