Ibutilide, a class III antiarrhythmic drug, has been shown to convert atrial fibrillation to sinus rhythm more rapidly than procainamide or sotalol. Our objective was to compare the efficacy and safety of ibutilide and amiodarone in patients after cardiac surgery. Prospective, randomized, double-blinded study. Intensive care unit of a university hospital. Forty adults with an onset of atrial fibrillation within 3 hrs after admission. Before the administration of antiarrhythmic drugs, a 24-hr Holter electrocardiograph was attached. Patients in the ibutilide group received ibutilide 0.008 mg/kg body weight over 10 mins; treatment was repeated if atrial fibrillation or flutter persisted. If sinus rhythm was not achieved within 4 hrs, amiodarone 5 mg/kg was administered over 30 mins, followed by amiodarone 15 mg/kg over 24 hrs. Patients in the amiodarone group received amiodarone 5 mg/kg over 30 mins, followed by amiodarone 15 mg/kg over 24 hrs if atrial fibrillation or flutter continued. Within the first 4 hrs, atrial fibrillation was converted in nine of 20 patients (45%) in group ibutilide and in ten of 20 patients (50%) in group amiodarone (not significant). Mean time for conversion overall was 385 mins in group ibutilide and 495 mins in group amiodarone (not significant). In group amiodarone, the protocol was discontinued in two patients because of severe arterial hypotension. Atrial fibrillation recurred in 11 of 20 patients (55%) in group ibutilide and in seven of 20 patients (35%) in group amiodarone (not significant). Ventricular arrhythmia did not occur during the first 24 hrs of the protocol. Ibutilide has no significant advantage over amiodarone for the conversion of atrial fibrillation to sinus rhythm in either time to conversion or conversion overall, but severe hypotension was not seen with ibutilide.