PURPOSE: To evaluate the effect of long-term treatment with metoprolol after coronary bypass grafting on death and cardiac events. METHODS: Patients in western Sweden on whom coronary artery bypass grafting was performed between June 1988 and June 1991 were evaluated for inclusion during the first 3 weeks after surgery. Major exclusion criteria were age > 75 years, concomitant valve surgery, traditional contraindications to beta-blockers and unwillingness to participate. Patients were randomized in a double-blind fashion to 100 mg of metoprolol/placebo daily for 2 weeks and thereafter 200 mg daily for 2 years. RESULTS: Of 2365 patients who were operated on, 967 were randomized to either metoprolol (n = 480) or placebo (n = 487). Primary end points (death, non-fatal myocardial infarction, unstable angina pectoris, need for coronary artery bypass grafting or percutaneous transluminal angioplasty), were reached by 42 patients in the metoprolol group (8.8%), as compared with 39 in the placebo group (8.0%) (P = 0.73). Of all the patients randomized to metoprolol, 34% withdrew from blind treatment prematurely compared with 44% for placebo (P < 0.01). CONCLUSION: Prophylactic treatment with metoprolol over a 2-year period after coronary artery bypass grafting did not reduce death or the development of cardiac events. However, the 95% confidence limits ranged from the possibility of a 30% reduction in events to a 68% increase in events if patients were treated with metoprolol as compared with placebo.