The aim of this prospective randomized study was to compare the efficacy and safety of catheter ablation (CA) versus surgical ablation (SA) in patients with paroxysmal (P) and persistent (Pers) AF after failed initial pulmonary vein isolation. The patients with symptomatic AF (59 PAF and 41% Pers AF) after a previously failed primary RF ablation procedure were eligible for this study. The patients were randomized to CA (n = 32) or video-assisted SA (n = 32) redo ablation. The primary end-point of the study was the absence of any atrial tachyarrhythmias after a second ablation procedure during 12-month follow-up. At the end of observation, 26 (81 %) of the 32 SA group patients were AF/AT-free. In contrast, in the CA group, only 15 (47%) of 32 patients were AF/AT-free (p=0.004, log-rank test). Both groups received no antiarrhythmic drugs. In patients with PAF, 17 (85%) patients of 20 in the SA group and 10 (56%) patients of 18 in the CA group were AF-free (p=0.04, log-rank test). In patients with Pers AF, 9 (75%) patients of 12 in the SA group and 5 (36%) patients of 14 in the CA group were AF-free (p=0.04, log-rank test). The total number of postoperative complications was higher in the SA group as compared with that in the CA group (7 and 1 complications respectively, р = 0.02). In patients with PAF and PersAF after failed initial catheter ablation, video-assisted thoracoscopic surgical ablation is superior to endocardial catheter ablation for maintenance of the sinus rhythm. However, SA results in a higher rate of complications.