The present study in hypertensive men (40-64 years old) with untreated diastolic blood pressure above 100 mm Hg was aimed at investigating whether metoprolol (n = 1,609) given as initial treatment would lower the risk for coronary events (sudden death and myocardial infarction) more effectively than thiazide diuretics (n = 1,625). A substantial part of this study was the metoprolol arm of the Heart Attack Primary Prevention in Hypertension (HAPPHY) study. The HAPPHY study was a pooling of the effect of different beta-blockers, mainly metoprolol and atenolol, in which no favorable effect in relative risk was observed for atenolol as compared with diuretics. In the present study, 255 patients suffered definite coronary events during follow-up; 25% of these events were fatal, 39% were acute myocardial infarctions, and 36% were silent myocardial infarctions. The risk for coronary events was significantly lower in patients on metoprolol than in patients on diuretics (111 versus 144 cases, p = 0.001, corresponding to 14.3 versus 18.8 cases/1,000 patient years and a relative risk of 0.76 at the end of the trial; 95% confidence interval 0.58-0.98). This difference in risk has potentially important implications for clinical practice because of the large number of hypertensive patients who are at increased risk for coronary events. Because a placebo group, for ethical reasons, could not be included, relative risk can only be expressed in relation to diuretics. There was no difference between the two treatment groups in baseline characteristics, blood pressure during follow-up, or stroke rates. Thus, the difference in risk for coronary events is probably mediated via mechanisms other than blood pressure control. However, present data might suggest that different beta-blockers may have different efficacy in preventing coronary events. The reasons for this possibility are as yet unknown.