Background: Complete occlusion of the pulmonary veins (PVs) with the cryoballoon (CB) is considered to be the crucial factor for a successful PV isolation (PVI). We investigated whether a complete occlusion was indispensable for a successful CB based PVI of every PV. Methods and Results: Atrial fibrillation patients (n=123, 97; paroxysmal) undergoing a de novo PVI were enrolled. A total of 477 PVs were analyzed. The occlusion grade (OG) was scored as follows: OG3 (complete occlusion), OG2 (incomplete occlusion with slight leakage), OG1 (poor occlusion with massive leakage). There was no significant difference in the CB temperature (CBT) at all measured time points (from 30 to 120sec after freezing) and nadir CBT between OG2 and OG3 in all PVs except for the right inferior PV (RIPV). The RIPV isolation success rate was significantly lower for the OG2 status than OG3 (97.5 vs. 57.6%; p<0.0001). In contrast, there was not significant difference in the isolation success rate of the other three PVs between OG2 and OG3. In particular, the success rate of the right superior PV (RSPV) isolation was 95% for both OG2 and OG3. Phrenic nerve paralysis (PNP) was provoked during the RSPV isolation in two patients in whom the RSPVs were frozen during OG3. Conclusions: An OG3 may not always be required for a successful PVI of all PVs except the RIPV. OG2 could have comparable effects as OG3 in terms of a successful RSPV isolation. Not aiming for OG3 for the RSPV may reduce the risk of PNP.