Abstract Background Pulmonary vein isolation (PVI) with cryobaloon ablation (CBA) and radiofrequency ablation (RFA) lead to a similar outcome of freedom from atrial tachyarrhythmia (ATA) during follow-up as showed by the NO-PERSAF study. Analyses of ATA recurrence for these two techniques based on data of electroanatomical mapping (EAM) are scarce from randomized clinical trials. Purpose This study aimed to evaluate and compare clinical outcomes and EAM data recorded in the left atrium (LA) during CBA and RFA procedures in patients with persistent atrial fibrillation (AF). Methods For the NO-PERSAF study, 101 patients (mean 63.2±8.6 years, 80 male, 21 female) with persistent AF (77 persistent AF, 24 long-standing persistent AF) referred to PVI were randomized (1:1) to CBA or RFA and followed for 12 months. A prolonged 3-year follow-up was carried out. EAM was performed by a circular mapping catheter with a 3-dimensional mapping system in all patients. EAM data were recorded and analyzed off-line. The areas of the LA, posterior wall, low-voltage (<0.5mV) and complex fractionated electrogram (cycle length <120 ms) were measured and corrected by body surface area (BSA). EAM data from 90 patients were qualified for further analysis and the others (5 in CBA and 6 in RFA) excluded because of disqualification or technical failure of recording. Correlation between clinical results of 3-year follow-up and the EAM data were analyzed using univariate cox regression. Results After 3-year follow-up, freedom from ATA was achieved in 51 (50.5%) of the 101 patients, 28 (54.9%) out of 51 patients in the CBA group and 23 (46.0%) out of 50 patients in the RFA group and no difference was found (p=0.327). RFA compared with CBA resulted in larger ablation lesions in the LA (23.9% vs. 21.3%; p=0.037) and particularly in the posterior wall (44.2% vs 31.8%; p<0.001). Compared between the patients with and without recurrence, there was no difference with respect to age, gender, AF history, body mass index, heart disease and CHA2DS2-vasc score. For EAM data, the area of the LA and low voltage were significantly higher in ATA recurrent patients in CBA group, but not in RFA group (Table 1). Univariate cox regression analyses suggested that 3-year follow-up outcomes were associated with the size of the LA and the amount of low voltage corrected by BSA. The criteria of the LA area >55 cm2/m2 or low voltage area <19 cm2/m2 could predict ATA recurrence after CBA, but not after RFA procedure (Figure 1). Conclusions The size of the LA and the amount of low voltage corrected by BSA was associated with the outcomes of PVI during follow-up and could be employed as predictors of ATA recurrence for CBA, but not for RFA procedure.Table 1Figure 1K-M curves after CBA and RFA