Transient sinus node dysfunction has been demonstrated by noninvasive methods following the maze procedure for atrial fibrillation (AF). However, extensive data from invasive electrophysiological studies have not been previously reported. Thirty-seven patients, mean age 54 +/- 10 years, underwent the maze (III) procedure. Electrophysiological studies with recordings of SNRT, CSNRT, AVN-ERP, point of Wenckebach block, AH, PA, and HV interval, were performed preoperatively and 6 and 15 months postoperatively. Induction of atrial flutter/AF was attempted postoperatively. Based on electrophysiological study evaluation, the maze (III) procedure did not cause permanent damage to the sinus node in any patient with a documented normal sinus node function preoperatively (CSNRT max 541 +/- 210 vs 587 +/- 437 ms, P = 0.26). Postoperative AV node function was normal in all patients with a documented normal AV node function before surgery. One patient had an iatrogenic third degree AV block. There was no difference in PA or HV interval after surgery. Sustained atrial tachyarrhythmias could be induced in 5 patients, of whom 4 developed permanent AF/atrial flutter late after surgery. At late follow-up, (mean 45 months), 27 (73%) patients were in sinus rhythm, 5 (13%) patients had permanent pacing, and 5 patients had recurrent AF requiring His bundle ablation (n = 2) or medical treatment (n = 3). Based on electrophysiological studies, the maze (III) procedure does not cause permanent damage to the sinus or AV nodes or to the right atrial and His-Purkinje conduction systems. Electrophysiological study evaluation may predict the need for postoperative pacemaker. Induction attempts of atrial arrhythmias may predict future recurrences and guide therapeutic efforts.