The effects of abrupt withdrawal of atenolol, a long acting cardioselective beta blocker, were studied in 20 patients with severe stable angina pectoris admitted to hospital for coronary arteriography. During the 144 hour postwithdrawal period no serious coronary events occurred. Mean and maximal daily heart rates rose steadily for at least 120 hours. No important arrhythmias were noted on ambulatory electrocardiographic monitoring. Treadmill exercise testing at 120 hours showed little reduction in the times to angina, ST depression, and maximal exercise when compared with those recorded at 24 hours. This deterioration was small when contrasted with the improvements in these indices produced by atenolol treatment in a similar group of patients not admitted to hospital. No change in catecholamine concentrations or acceleration of the heart rate response to exercise occurred after atenolol withdrawal, suggesting that rebound adrenergic stimulation or hypersensitivity was absent or insignificant. Catastrophic coronary events after beta blockade withdrawal (the beta blockade withdrawal syndrome) have occurred almost exclusively in patients taking propranolol, many of whom had unstable angina at the time of withdrawal. This study showed that in patients with stable angina, even when severe, the abrupt withdrawal of atenolol can be expected to result in only minor clinical consequences. The risk to any patient of so called rebound events after withdrawal of beta blockade seems to be related to both the clinical setting and the agent being used.