Back Ground and Objectives Increased left ventricular mass (LVM) in patients undergoing aortic valve replacement(AVR) is associated with excess cardiac mortality and morbidity. The aim of the study is to investigate the determinant factors on LVM regression following AVR for pure aortic stenosis(AS). Method and Results In 193 patients undergoing AVR for AS, clinical and hemodynamic data were collected from their medical record retrospectively. Exclusion criteria included significant associated valvular disease to acquire pure AS patients. Echocardiography was performed preoperatively, early (7짹3) days after AVR ('early') and (11짹1) months (���late���). After AVR, the LVMI regression was significant both at 'early' (142짹46 vs. 157짹46 g/m2, regression against preoperative value, P<0.001) and ���late��� (112짹41 vs. 142짹46, against early value, P<0.001) evaluations. Multivariate analyses revealed that preoperative LVMI (P<0.001), preoperative transvalvular mean pressure gradient(mPG) of aortic valve (P=0.02) and mPG difference between the preoperative and the ���late��� (P=0.004) showed significant influence on LVMI change, but age, sex, valve type, valve size, preoperative ejection fraction, diastolic LV internal diameter, relative wall thickness and atrial fibrillation had no significant influence. Conclusions LVM regression is significant both early 7 days and one year after AVR. Preoperative LVMI, preoperative transvalvular mPG of aortic valve and mPG change for one year after AVR are determinant factors on LVM regression following AVR for pure AS.