Background: Areas of complex fractionated electrograms (CFAEs) have been reported to represent critical substrate of the atrial fibrillation (AF), which could be one of effective target of catheter ablation of AF. However, incomplete CFAE ablation may have a potential to develop atrial tachyarrhythmias. The purpose of this study is to compare the effect & clinical outcome between CFAE(s)-guided linear ablation and anatomically guided linear ablation in the patients with persistent AF. Methods: Of 238 consecutive patients with persistent AF, 22 patients who were performed CFAE(s)-targeted linear ablation (group 1) and 92 patients who underwent anatomically pre-determined linear ablation (group 2) after circumferential isolation of 4 pulmonary veins (PV) were compared. Results: Baseline characteristics including age, sex, LA size, and LVEF were not significantly different between two groups. Procedure time was longer in group 1 than in group 2 (335.0짹75.1 vs. 285.5짹79.6 minutes, p=0.011), however, ablation time was not significantly different (171.6짹51.7 vs. 145.6짹56.7 minutes, p=0.091). After the procedure, inducibility of AF or atrial tachycardia was not different between two groups (22.7% in group 1 vs. 13.2% in group 2, p=0.261). During the mean 25.9짹16.2 months, the rate of freedom from recurrence was significantly higher in group 1 compared with group 2 (13.6% vs. 38.5%, p=0.042). In the Kaplan-Meier analysis, there was no significance difference in the recurrence between two groups. Procedure related serious complications were noted in only group 2 (cardiac tamponade 4, stroke 2). Conclusion: In addition to PV isolation, CFAE(s)-guided linear ablation is more effective and safe compared to anatomically determined linear ablation in patients with persistent AF. Prospective study with longer follow up periods is required.