The effect of successful catheter ablation of overt accessory pathways on the incidence of atrial fibrillation (AF) was studied in 129 symptomatic patients with (n = 75) or without (n = 54) previous documented AF. Fourteen had had ventricular fibrillation. Factors predictive of recurrence were examined, including electrophysiologic parameters. Atrial vulnerability was defined as induction of sustained AF (greater than 1 minute) using single, then double, atrial extrastimuli at 2 basic pacing cycle lengths. When compared to patients with only reciprocating tachycardia, patients with clinical AF included more men (77 vs 54%, p = 0.008) and were older (35 +/- 12 vs 29 +/- 12 years, p = 0.01). They had a significantly shorter cycle length leading to anterograde accessory pathway block (252 +/- 42 vs 298 +/- 83 ms, p less than 0.001), greater incidences of atrial vulnerability (89 vs 24%, p less than 0.001) and subsequent need for cardioversion (51 vs 15%, p less than 0.001). After discharge, the follow-up period was 35 +/- 12 months (range 18 to 76); 7 patients with previous spontaneous AF (9%) had recurrence at a mean of 10 months after ablation. Age, presence of structural heart disease accessory pathway location, atrial refractory periods and accessory pathway anterograde conduction parameters were not predictive of AF recurrence. Persistence of atrial vulnerability after ablation was the only factor associated with further recurrence of AF. Atrial vulnerability was observed after ablation in only 56% of patients with previous AF versus 89% before ablation. It is concluded that successful catheter ablation of accessory pathways prevents further recurrence of AF in 91% of patients.(ABSTRACT TRUNCATED AT 250 WORDS)