Background. Atrial fibrillation is a common complication of cardiovascular surgery. β-Blockers have been shown to decrease the incidence of postoperative atrial fibrillation. However, the use of magnesium is more controversial. It was our hypothesis that adjunctive magnesium sulfate would improve the efficacy of β- blockers alone in the prevention of postoperative atrial fibrillation. Methods. We prospectively randomized 167 coronary artery bypass patients (mean age 61 ± 10 years, 115 men) to receive propranolol alone (20 mg four times daily) or propranolol and magnesium (18 g over 24 hours). Magnesium was begun intraoperatively, and propranolol was started on admission to the intensive care unit. Results. Using an intention-to-treat analysis, the incidence of postoperative atrial fibrillation was 19.5% in the propranolol-treated patients and 22.4% in propranolol + magnesium- treated patients (p = 0.65). Because combination therapy resulted in an excess of postoperative hypotension, which required withholding doses of propranolol, an on- treatment analysis was also performed. In this analysis, the incidence of atrial fibrillation was still not significantly different (18.5% in propranolol-treated patients and 10.0% in propranolol + magnesium-treated patients, p = 0.20). Conclusions. Adjunctive magnesium sulfate, in combination with propranolol, does not decrease the incidence of postoperative atrial fibrillation. (C) 2000 by The Society of Thoracic Surgeons.