Myocardial infarction, sudden cardiac death, cerebrovascular accidents, blood pressure control and treatment tolerability were studied in a randomized double-blind trial conducted in 6357 men and women aged 40-64 years with uncomplicated essential hypertension (diastolic blood pressures 100-125 mmHg). At the start of the trial 3185 patients received treatment based on a beta-blocker (oxprenolol), while in the remaining 3172 placebo replaced oxprenolol. Supplementary drugs, excluding beta-blockers, were used as necessary in both treatment groups, with the aim of reducing diastolic pressure to 95 mmHg or less. Patients were followed for 3-5 years, a total of 25 651 patient years at risk. In most respects the two groups fared equally well; sudden death (relative risk [RR] 1.08; 95% confidence interval [Cl] 0.68 and 1.72), myocardial infarction (RR 0.83; Cl 0.59 and 1.16) and cerebrovascular accident (RR 0.97; Cl 0.64 and 1.47) rates were similar. Beta-blocker based therapy was associated with significantly lower average blood pressures, earlier ECG normalization, less hypokalaemia and fewer withdrawals from double-blind treatment for uncontrolled hypertension. Doctor-elicited and patient-assessed unwanted effects demonstrated overall good tolerability. In smokers the cardiac event rate was doubled. We propose that beta-blocker treatment effects depend on smoking status, with a significant interaction benefiting non-smoking men. Lower blood pressures during treatment were associated with substantially lower rates for cardiac as well as cerebrovascular events. Proportional hazards analysis also underlines the importance of other cardiovascular risk factors. The IPPPSH stresses the need for a comprehensive approach to the management of blood pressure and other risk factors in hypertensive patients.