The factors that make mitral reconstruction an attractive alternative to valve replacement are increased anatomical valve orifice, preservation of the valvular apparatus, improved longevity over porcine xenografts, and no requirement for long-term anticoagulation. In our experience the majority of patients with mitral regurgitation have degenerative valve disease. This comparative series of patients operated on over a 10 year interval includes 72 with replacement and 112 with reconstruction. The average cross-clamp times were 69 minutes for reconstruction and 44 minutes for replacement. The hospital mortalities were 3.6% and 18.1%, respectively. Postoperative valve areas as determined by Doppler echocardiography were 2.69 cm2 for replacement and 2.67 cm2 for reconstruction (p = 0.9). The valve failure rate was 2.0% per patient-year for reconstruction and there were no failures in the replacement group. The reoperation incidence for reconstruction was 20% at 10 years. The remaining patients, although clinically well, had varying degrees of stable, nonprogressive mitral regurgitation. Forty-five percent had mild to trivial regurgitation detectable by echocardiographic studies. The decision between mitral valve reconstruction and mitral valve replacement remains a highly controversial issue. The failure of our data to demonstrate superior valve function for patients with mitral regurgitation undergoing valve reconstruction suggests a need for careful analysis of reconstruction with respect to effective valve orifice and incidence of regurgitation.