Background The objective of this study was to assess the probability of development of heart failure during a long-term follow-up in patients submitted for aortic valve replacement for aortic regurgitation on the basis of preoperative findings. Methods and Results Eighty-seven consecutive patients with pure aortic regurgitation and normal coronary arteries were submitted for aortic valve replacement and prospectively followed up. Clinical examination, echocardiography, and radionuclide ejection fraction were performed before surgery and at 1, 2, 5, and 10 years after surgery. Operative mortality rate was 2.2% (2 patients). The follow-up period was 1 to 12 years (mean 6 years). Overall survival rate was 87% at 5 years and 81% at 10 years. During follow-up, 19 patients had heart failure develop, and there were 14 deaths (6 caused by heart failure). Probability of heart failure was 16% at 5 years and 24% at 10 years. Age was the single independent preoperative predictor of both death and heart failure. Age >50 years (relative risk [RR] 10.4), preoperative ejection fraction <40% (RR 10.6), and end-systolic diameter >50 mm (RR 7.4) were independently related to the postoperative development of heart failure. Conclusions Aortic valve replacement can be performed safely in patients with severe aortic regurgitation by following current recommendations. Age >50, end-systolic diameter >50 mm, and radionuclide ejection fraction <40% were independent preoperative predictors of postoperative heart failure. The only independent predictor of both postoperative death and heart failure was age >50 years. (Am Heart J 1998;136:681-7.)