Propranolol and practolol were tested in patients with repeated daily occurrence of spontaneous angina. Twenty-one showed ST segment depression (type I) and 15 ST segment elevation (type II) during angina. The efficacy of the treatment was evaluated in subjective (number of reported episodes of pain) and objective terms (number of episodes of electrocardiographic abnormalities documented during periods of continuous recording): practolol was fully effective in 42 per cent and propranolol in 38 per cent of type I cases; in type II angina 73 per cent of the cases fully responded to propranolol, none of the patients in this group given practolol improved. The study also showed that: (a) the effects on angina are strictly dose-dependent, and optimal results are achieved at individualized doses; (b) within the same subject the response may be preferential to one beta-blocker as opposed to the other; (c) propranolol is more effective in type II angina; (d) the occurrence of heart failure is uncommon even with high doses of beta blockers;(e) the relief of angina is due to prevention of ischaemia and not to a placebo or anaesthetic effect; (f) the prevention of ischaemia is not adequately explained by reduction of the mechanical effort and the oxygen need of the myocardium; (g) the antianginal effect is possibly dissociated from the beta blockade of the heart. The hypothesis that beta-blocking agents influence the conronary vasomotion is discussed.