The beta-adrenergic selectivity of diacetolol, the major metabolite of acebutolol, has been compared with that of acebutolol, metoprolol and propranolol in 11 normal subjects. Bronchial and cardiac beta-adrenoceptor blockade were assessed on separate occasions after diacetolol 600 mg, acebutolol 400 mg, metoprolol 200 mg, propranolol 80 mg and placebo. Bronchial beta-adrenoceptor blockade was assessed as the displacement of the bronchodilator dose response curve to inhaled isoprenaline after each beta blocking drug compared to placebo and expressed as the dose ratio. Bronchodilatation was measured as change in specific airway conductance (sGaw) in the body plethysmograph. Cardiac beta-adrenoceptor blockade was assessed as the percentage reduction in exercise heart rate during the 5th minute of exercise at 70% of the subject's maximum work rate. There was a significant reduction in exercise heart rate with all 4 beta-blocking drugs when compared with placebo, 22% for diacetolol, 24% for acebutolol, 25% for propranolol and 28% for metoprolol. The reduction with metoprolol was significantly greater than the reduction with the other three beta-adrenoceptor antagonists. Mean dose ratios for the airway isoprenaline dose response curves after each of the 4 beta-blocking drugs were 2.4 for diacetolol, 2.7 for metoprolol, 8 for acebutolol and 72 for propranolol. The difference between diacetolol and metoprolol was not significant. Thus diacetolol appears to be more cardioselective than acebutolol and both are more cardioselective than propranolol in man. Metoprolol is probably more cardioselective than diacetolol though interpretation of the differences in exercise heart rate is complicated by the fact that diacetolol has some intrinsic sympathomimetic activity.