Ablation index (AI) is a novel technology of ablation lesion quality to help improve homogeneity of lesion size and continuity. In this study, we aim to evaluate whether AI-guided PVI improves clinical outcomes compared to CF-guided PVI in patients with paroxysmal AF (PAF).Patients undergoing first-time radiofrequency ablation for PAF were randomized in a 2:1 ratio to two groups: AI-guided PVI and CF-guided PVI. In the AI group, AI ≥500 was recommended at the anterior/superior/inferior walls, 350-400 at the posterior wall, and inter-lesion distance ≤4 mm. The primary endpoint is the freedom from atrial arrhythmia recurrence during 12 months follow-up, without antiarrhythmic drug therapy (ADT). The key secondary endpoints include intra-procedural efficiency and peri-procedural complications.225 patients were randomized (AI group (n = 149) and CF group(n = 76)). First-pass isolation rate in AI group was significantly higher than that in CF group (58.3% vs. 43.4%, p = 0.035). After a median follow-up of 12.2 months, 154/225(68.4%) of patients were free from atrial arrhythmia recurrence without ADT, which was higher in AI group compared with CF group, but without significant difference (71.1% vs. 63.2%, p = 0.253). The incidence of peri-procedural complications is low and without difference between two groups.AI-guided ablation provided higher acute efficacy than CF-guided ablation in PV isolation for patients with paroxysmal AF. The long-term success rate in AI group was higher than CF group, but did not reach statistical significance. This article is protected by copyright. All rights reserved.This article is protected by copyright. All rights reserved.