Objectives To determine the learning curve of remote magnetic navigation (RMN)-guided ablation for idiopathic ventricular arrhythmias (IVA) and its implications on the outcomes of the procedures. Methods A total of 249 IVA patients undergoing catheter ablation with RMN were divided into two groups: right ventricle (RV) and left ventricle (LV) group regarding to the initially mapping from RV or LV. Procedural outcomes and complications were compared between two groups. The learning curve was determined by plotting procedure time of the consequentially performed procedures in RV group patients. These patients were then divided equally into 3 phases (P1, P2, P3) to determine the effect of learning curve on procedure outcomes. Results The acute success rate reached 91% and complication rate was only 0.4%. The procedure time, mapping time, ablation time and X-ray time of LV group was significantly higher than RV group (P<0.05), but no significant difference was detected in the complication rate of the two groups(P=0.313 3). The procedure time of RV group patients decreased significantly along the learning curve. Compared with P2 and P1, the procedure time of P3 was decreased by 24% (P<0.01) and 41% (P<0.000 1) respectively. The X-ray time was relatively low [(3.8±4.1) min] at the very beginning and dropped significantly along the learning curve [(1.3±1.0) min, P<0.001]. Conclusions It takes rather short term to establish the learning curve of RMN-guided IVA with low rate of complication. The radiation exposure for both patient and operator is low during the establishment of learning curve and could be further reduced later on.