OBJECTIVE At present, surgery is the only recommended effective treatment for severe aortic stenosis. However, the surgical risk is increased when left ventricular dysfunction is present. The aim of this study was to identify predictors of postoperative and long-term mortality and functional improvement after valve replacement in patients with severe aortic stenosis and left ventricular dysfunction. METHODS Between 1996 and 2008, 635 consecutive patients with severe aortic stenosis underwent surgery. Early postoperative mortality in the 82 with an ejection fraction <40% was 19.5%. The following independent predictors of early postoperative mortality were identified: female sex (odds ratio [OR]=2.60; 95% confidence interval [CI], 2.20-89.0; P=.004), mild mitral regurgitation (OR=2.38; 95% CI, 1.40-80.0; P=.020) and coronary artery disease (OR=2.09; 95% CI, 1.26-51.0; P=.027). RESULTS During the mean follow-up period of 42.59+/-40.83 months, overall mortality was 18.8% and cardiovascular mortality was 11.3%. The only factor associated with increased mortality during follow-up was a low postoperative cardiac output (OR=4.40; 95% CI, 1.20-15.5; P=.02). In total, 70.5% showed early improvement in ventricular function, the predictors of which were: no improvement following a previous myocardial infarction (P=.04), no revascularized coronary lesions (P=.04), and a low aortic valve pressure gradient (P=.02). Functional class improved significantly during follow-up in 93.4% of patients. CONCLUSIONS Despite considerable early postoperative mortality in patients with aortic stenosis and left ventricular dysfunction, over the long term there was evidence of better survival coupled to improved ventricular function and functional class.