Background The efficacy of mere pulmonary vein isolation epicardially for the treatment of permanent chronic atrial fibrillation, in comparison with the left atrial endocardial maze procedure was evaluated. Methods Retrospective data collection and analysis toward the outcome of 72 consecutive patients who underwent left atrial maze procedures between January 2003 and December 2005 was performed. Surgical ablation was performed concomitantly with valve and (or) coronary procedures. Group I (n = 29) received an endocardial left atrial ablation using unipolar saline irrigated radiofrequency (Medtronic Cardioblate surgical ablation pen; Medtronic Inc, Minneapolis, MN). Group II (n = 43) received epicardial isolation of the pulmonary veins using bipolar saline irrigated radiofrequency (Medtronic Cardioblate). Follow-up included 24h electrocardiogram and echocardiography 6 and 12 months postoperatively. Results Mean follow-up was 19.5 ± 1.0 months (17.7 ± 19.5 months group I vs 20.6 ± 1.1 months group II). Both groups were comparable with regard to duration of preoperative atrial fibrillation, European system for cardiac operative risk evaluation, left ventricular ejection fraction, aortic cross-clamp time, bypass time, intensive care unit and hospital stay ( p > 0.05). No maze procedure-related mortality was observed. In group I, three patients required postoperative pacemaker implantation due to atrioventricular (AV) bloc, bradycardia, and sick sinus syndrome, respectively. In group II, five patients required postoperative pacemaker implantation (three AV bloc and two bradycardia). Freedom from atrial fibrillation at last follow-up was 85.7% and 58.5% in groups I and II, respectively ( p = 0.016). Conclusions Pulmonary vein isolation alone seems to be insufficient in treating permanent chronic atrial fibrillation. In case of chronic permanent atrial fibrillation, left atrial endocardial maze, providing the connection lines to the mitral annulus and (or) between the pulmonary veins, seems to be mandatory.