Backgrounds; Atrial flutter (AFL) frequently occurs after isolation of pulmonary veins and/or left atrial substrate modification for atrial fibrillation (AF). However, their prevalence and clinical significance remain to be determined. Methods; 139 cases (re-do; n=25) in 114 consecutive patients (age 53.5 짹 13.3 years) for paroxysmal (n=85) or persistent (n=29) AF, who had frequently episodes of AF refractory to antiarrhythmic therapy, underwent AF ablation procedure by 3 different methods (PV potentials guided, n=44, trigger guided, n=39, and noncontact mapping guided, n=56). Spontaneous or induced AFL by rapid atrial pacing (RAP) after AF ablation was classified as typical or atypical AFL according to the isthmus-dependency. Results; Among 139 cases, typical (n=52) or atypical AFL (n=18) was documented in 56 patients (49.1%) and all of typical isthmus-dependent AFL was eliminated by cavo-tricuspid isthmus (CTI) block. Of 18 atypical AFL, either spontaneously (n=8) or by RAP (n=10), only two of them were eliminated by the enforcement of ablation on the left atrium or re-connected PV fascicle, and remaining 16 were converted to sinus rhythm by internal DC cardioversion. Among these 16 patients, 8 patients (50%) recurred AF (n=6) or AFL (n=2) during follow-up of 657짹506 days. There was no differences in the recurrence rate of AF/AFL among 3 methods for AF ablation (p=0.528). 86 patients for paroxysmal (71/85, 83.5%) or persistent (15/29, 51.7%) AF showed freedom from recurrent AF with (n=48) or without (n=38) drugs. Atypical AFL was significantly associated with recurrence of AF (p=0.011). Conclusions; AFL was frequently observed after AF ablation procedure and 80% of these episodes were typical AFL, which can be successfully eliminated by CTI block. Atypical AFL was significantly related to recurrence of AF rather than AFL during long term follow-up. Clinical efficacy of AF ablation could be improved targeting AFL after AF ablation.