Objective: The aim of this study was to assess the effects of aortic valve replacement (AVR) on the recovery of left ventricular function and the predictors for long-term survival in patients suffering from isolated severe aortic stenosis (AS) with a significant left ventricular dysfunction (LVD). Methods: This retrospective study was conducted on 46 patients with isolated severe AS and LVD [left ventricular ejection fraction (LVEF) ≤ 40%] who underwent AVR in our clinic between January 1993 and March 2006. Patients with coronary artery disease, with more than moderate aortic regurgitation (>2), with previous valve replacement or repair, and with other valve pathologies were excluded. The mean aortic valve area was 0.7±0.09 cm2. The following fourteen variables were analyzed: etiology, age (≥70 years), sex, preoperative New York Heart Association (NYHA) functional class, chronic obstructive pulmonary disease, hypertension, diabetes, peripheral arterial disease, chronic renal insufficiency, need for concomitant procedures for the ascending aorta, cardiopulmonary bypass time ≥120 min, aortic cross-clamp time ≥90 min, intraaortic balloon pump support and inotropic support. Statistical analysis for comparison of pre- and postoperative changes in clinical and functional variables was performed using Wilcoxon rank test. The predictors of early mortality after AVR were analyzed using logistic regression analysis and late survival was studied using Cox proportional regression and Kaplan Meier survival analyses. Results: Operative mortality was 8.6% with four patients. As the result of univariate logistic regression analysis, preoperative NYHA functional class ≥3 was found to be predictive of early mortality. Patients with NYHA class ≥3 had 12.6 times (OR: 12.6; 95%CI: 1.2-131.3; p=0.035) higher probability of early mortality than those with a lower NYHA class. However, multivariate logistic regression analysis demonstrated no predictor for early mortality. A positive change was observed in the LVEF in 79.3% of survivors and the mean LVEF increased from 34.5%±3.9% to 44.7%±10.4% (p<0.001). There were eight (19%) late deaths. Actuarial survival was 83.1% ± 5.9% at 5 years and 59.6%±10.9% at 10 years. Cox proportional hazards regression analysis demonstrated diabetes mellitus (p=0.031; HR: 6.6; 95% CI: 1.19-36.9) and intraaortic balloon pump use (p<0.001; HR: 10.7; 95% CI: 2.9-39.7) as significant predictors for late mortality. Conclusion: Left ventricular ejection fraction and symptoms improve after AVR in patients with isolated severe AS and LVD with an acceptable operative mortality and satisfactory long-term survival.