Background: Ablation index (AI) linearly correlates with lesion depth and may yield better therapeutic performance in pulmonary vein isolation (PVI) when tailored to a patient’s wall thickness (WT) in the left atrium (LA). Methods and results: (First study) In paroxysmal atrial fibrillation patients (PAF, n=20), the average LA WT (mm) in each anatomical segment for PVI was measured by intra-cardiac echocardiography (ICE) placed in the LA; the optimal AI for creating one-millimeter transmural lesion (AI/mm) was calculated. (Second study) PAF (n=80) patients were randomly assigned either to a force-time integral protocol (FTI, 400 gram·second, n=40) or a tailored-AI protocol (TAI, n=40). In TAI, the LA WT in each segment was individually measured by ICE before starting ablation; a target AI was adjusted according to the individual WT in each segment (AI/mm×WT). The acute procedure outcomes and the 1-year AF recurrence rate were compared between FTI and TAI. TAI had higher success rate of first-pass isolation and had lower incidence of residual PV-potentials/conduction gaps after a circular ablation than FTI (88% vs. 65%, 15 vs. 45%, respectively). The procedure time to complete PVI decreased in TAI compared to FTI (52 vs. 83 minutes), being attributed to the increased radiofrequency power and the decreased radiofrequency application time in each point in TAI. TAI had lower 1-year AF recurrence rate than FTI. Conclusion: WT-based AI-adjustment increased acute procedure success, decreased time for PVI, and reduced 1-year AF recurrence rate. Understanding the precise ablation target would improve the efficacy of PVI.