Backgrounds: Although left atrial emptying fraction (LAEF) can be precisely assessed by multislice computed tomography (MSCT), it is a time consuming technique because of requiring manual tracing the endocardial contour of the LA of each cross sectional image. The purpose of this study is to compare LAEF by conventional method with LAEF by the simple estimation method using LA volume measured by 3 orthogonal LA dimensions. Methods and Results: MSCT was performed in patients with paroxysmal (PAF, n=38) and non-paroxysmal (non-PAF, n=30) atrial fibrillation (AF) undergoing catheter ablation. Axial multiphase images were reconstructed using retrospective electrocardiogram gating. Images were reconstructed at 10 phases of the cardiac cycle. LA volume was measured at three points: (1) at maximal LA volume, (2) just prior to onset of atrial contraction (MDV, Mid-diastolic volume), and (3) at minimal LA volume using conventional and simple methods. LAEF was calculated from LA volumes measured; LAEFTotal = (LAVMax - LAVmin) / LAVMax * 100 (%), LAEFPassive = (LAVMax - LAVMDV) / LAVMax * 100 (%), LAEFActive = (LAVMDV - LAVmin)/ LAVMDV * 100 (%). During the sinus rhythm, there were significant correlations in LAEF between two methods (LAEFTotal 37.5짹10.5 % vs. 34.9짹13.7 %, r=0.711, p<0.001, LAEFPassive 18.8짹7.8 % vs. 9.6짹7.2 %, r=0.652, p<0.001, LAEFActive 23.0짹10.6 % vs. 28.1짹13.8 %, r=0.515, p=0.001). During the AF, LAEFTotal measured by simple method was closely correlated with that by conventional method. (LAEFTotal 15.4짹6.9 % vs. 13.3짹8.9 %, r=0.602, p=0.002). Conclusion: Irrespective of underlying rhythm, LAEF measured by simple method was strongly correlated with that obtained by conventional method. Simple method was a feasible and alternative method to estimate LAEF in patients with AF.