We studied the efficacy of propafenone in preventing atrial tachyarrhythmias after cardiac surgery, and the possible relationships between CYP2D6 polymorphism and the efficacy, pharmacokinetics, and tolerability of propafenone. One hundred and sixty patients were randomized (double blind) to receive propafenone (n= 78) or placebo (n= 82) for 1 week after cardiac surgery. The patients who were assigned to the propafenone group received 1 mg/kg infused in 1 h, followed by a continuous infusion at a rate of 4 mg/kg/24 h until the following morning, and subsequently 450 mg/day orally until the sixth postoperative day. Thirty-seven patients completed the trial in the propafenone group and 45 in the placebo group. The frequency of occurrence of atrial tachyarrhythmia was lower in the propafenone group than in the placebo group (29.7% vs. 53.3%, P< 0.05; relative risk, 0.56). Plasma propafenone concentrations were markedly influenced by CYP2D6 genotype-derived phenotype.