Background: Integration of electroanatomic mapping and 3DCT(CARTO-Merge) has been gaining acceptance to facilitate catheter ablation of atrial fibrillation(AF). We tried to evaluate the practical accuracy, and to find out the common causes and factors affecting the errors. Material and Methods :70 patients undergoing catheter based AF ablation with CARTO-Merge software were enrolled. The practical match accuracy was graded with 5 points(5:excellent to 1:very bad). We invested the major and minor errors interfering the procedure and the causes of each errors. And the statistical analysis were done with match statistic data provided by the software and practical match accuracy, and other factors affecting errors. Results:In 70(45 men,mean age:59짹10) patients, 23(32.9%) patients showed major and minor non-correctable match errors. The causes were technical match errors(10, 43%), position change due to movement(5, 21.7%), volume discrepancy with CT and 3D map(4, 17.4%), rotational disarrangement(3, 13.0%), and bad CT image(1, 4.3%). It showed significant correlation between LA size and match statistic data (r=0.382, p=0.01). Excluding patients movement, match statistic data showed significant correlation with LA size (r=0.382, p=0.002) and practical match accuracy (r=-0.276, p=0.026). Conclusion :Although CARTO-Merge is very useful tool for AF ablation, to have accurate integration, we need to do more detailed mapping, try to prevent the movement, and consider the position of patients, CT image and LA size.