Objective: The authors analyzed the clinical results during the first 8 1/4 years' experience with the Maze procedure for the surgical treatment of atrial fibrillation. Summary Background Data: Atrial fibrillation occurs in 0.4% to 2% of the general population and in approximately 10% of patients older than 60 years of age. It is associated with significant morbidity and mortality. The irregular heartbeat causes discomfort, the loss of synchronous atrioventricular contraction compromises hemodynamics, and the stasis of blood flow increases the vulnerability to thromboembolism. Methods: From September 25, 1987 to March 1, 1996, 178 patients underwent the Maze procedure. Thirty two patients underwent the Maze-I procedure, 15 underwent the Maze-II procedure, and 118 underwent the Maze-III procedure. Patients were analyzed for recurrence of atrial flutter and atrial fibrillation between 3 months and 8 1/4 years after surgery (n = 164). Patients were analyzed for atrial transport function, sinus nodule function, and postoperative pacemaker requirements. Results: Ninety three percent of all patients were arrhythmia free without any antiarrhythmic medication. Of the remaining patients with arrhythmia recurrence, all were converted to sinus rhythm with medical therapy. All patients were documented to have atrial transport function by either direct visualization, transesophageal echocardiography, or atrioventricular versus ventricular pacing at the same rate. Ninety eight percent had documented right atrial function, and 94% had left atrial function of the 107 patients in this series who were documented to have a normal sinus node preoperatively, only 1 patient required a permanent pacemaker. Conclusion: The Maze procedure is an effective treatment for medically refractory atrial fibrillation in properly selected patients.