Introduction: Targeted treatment of atrial fibrillation (AF) involves circumferential isolation of all pulmonary veins (PV) or isolation of electrical connections within their ostia. Only in some cases are the real localisation and number of triggering foci, the anatomy of venous ostia as well as the form of AF (paroxysmal, persistent, chronic, primary or secondary) taken into consideration. Aim: To compare the results of selective electrical isolation (1-3 PV ostia or ablation of a single focus in other veins or atrium) versus isolation of all pulmonary veins. Methods: RF ablation was performed in eighty patients (51 men, 29 women) with symptomatic, drug-refractory AF. Fifty-nine patients had paroxysmal AF (PAF), 16 persistent (AFpers), and 5 chronic AF (AFchro). Selective ablation was carried out in those patients who had detectable AF triggers during sinus rhythm – supraventricular extrasystolic beats (SVEB) of 1 to 3 morphologies (group I). Extended ablation – isolation of all 4-5 PV – was performed in patients with multiple SVEB morphologies and heterogeneous electrical connections within all PV (group II). Group I consisted of 60 patients (22 females) aged 46±14 years, whereas group II comprised 20 patients (7 females) aged 52±13 years. In 24 patients (18 from group I and 6 from group II) with concomitant typical atrial flutter, an ablation line in the cavo-tricuspid isthmus was also performed. Long-term results were assessed 17±15.6 (4-105) months after the procedure based on routine ECG, ambulatory 24-hour ECG monitoring, clinical evaluation and regular phone calls. In patients with PAF, left atrial diameter <4.2 cm and evidence of successful ablation, antiarrhythmic agents were withheld. In patients with AFpers and AFchro, antiarrhythmic drugs were discontinued 3 to 6 months after successful ablation. Results: Complete procedural success was achieved in 61 (76%) patients, and significant clinical improvement was observed in another 9 (11%) patients. Effective ablation significantly improved quality of life. In group I the procedure was entirely successful or a marked improvement was reported (single, transient palpitation episodes and/or atrial tachyarrhythmias lasting up to 30 seconds) in 54 (90%) patients. Among 48 (80%) patients with complete success, 25 (42%) did not receive any antiarrhythmic drugs during follow-up, 12 (20%) with arterial hypertension received β-blockers, and 11 (18%) continued β-blocker + class I antiarrhythmic drug. In another 6 (10%) patients a significant clinical improvement in arrhythmia control was observed. In Group II the procedure was fully effective or a significant improvement was observed in 16 (80%) patients. Among 13 (65%) patients with complete success, 5 (25%) did not require any antiarrhythmic drugs, 4 (20%) who had hypertension continued β-blockers, and another 4 (20%) continued β-blocker + I class antiarrhythmic drug. A significant clinical improvement of arrhythmia control was observed in another 3 (15%) patients. Conclusions: In patients with a limited number of triggering foci and limited AF substrate, selective ablation effectively eliminates AF with a low risk of complications. Detailed electrophysiological assessment (standard ECG, 12-lead Holter ECG monitoring and endocardial mapping) allows precise identification of this group of patients. In patients with chronic and persistent AF benefits occur with some delay which is associated with a delayed reversal of atrial remodelling