Objective:This study was designed to investigate prognosis differences among different negative induction test results in patients with paroxysmal atrial fibrillation underwent circumferential pulmonary vein isolation(CPVI). Methods:We retrospectively studied 133 patients (PTs)who underwent catheter ablation due to paroxysmal atrial fibrillation with the endpoint of non-inducibility(defined as atrial arrhythmias could not be induced or can be induced but lasted less than 3 minutes). The induction protocol was listed as follows:after successful CPVI,which was defined as completion of ablation set and bidirectional blockade of pulmonary vein(PV)-left atrium (LA)conduction,decremented burst stimulation (10 mA,2 ms pulse width)was attempted at coronary sinus orifice(CSo),and distal of coronary sinus (CSd)from 300 ms to loss of atrium capture (1∶1). If sustained (lasting >3 minutes)atrial arrhythmias were induced,the key site would be identified and further ablation performed. According to different negative induction test results at endpoint,patients were divided into group A (atrial arrhythmias could not be induced)and group B (atrial arrhythmias could be induced with duration< 3 minutes). All the patients were followed up regularly and receive free ECG/Holter. The blanking period was 3 months. Results:According to induction test results,74 PTs(55.6%) were detected in group A and 59 PTs(44.4%) in group B. After a mean follow-up of (21.3 ± 10.9)months,25 PTs in group A and 20 PTs in group B had AF relapse since the index ablation. Kaplan-Meier survival analysis showed no significant differences of AF relapse ratio between group A and B (P = 0.74). Conclusion:For patients with paroxysmal AF who underwent CPVI with the endpoint of non-inducibility,those who atrial arrhythmias could not be induced at the endpoint do not promise a better prognosis than those who have short(duration< 3 minutes)atrial arrhythmias could still be induced.