Patients with ischemic mitral incompetence have a high operative risk whether the valve is repaired or replaced. The advantage of repair over replacement is unclear in this group of patients.Between April 1986 and December 1994, 232 patients underwent surgery for ischemic mitral valve insufficiency; mitral valve replacement was performed in 98 of them. Operative mortality was 13.3%. The actuarial survival rate after 5 years was 73.3%. The surgical risk in patients whose left ventricular ejection fraction (LVEF) was 10%-30% (operative mortality 50.0%) was higher than in those whose LVEF was greater than 30%. Valve reconstruction was performed in 102 patients. Operative mortality in this patient group was 14.7%. The surgical risk in patients whose LVEF was < or = 30% was higher (operative mortality 42.9%).The total actuarial survival rate of all patients was 64.4% after 5 years. Mortality during follow-up was higher in patients with residual mitral valve insufficiency greater than grade I after mitral valve reconstruction. Twenty-four patients with severely impaired left ventricular function underwent heart transplantation. Operative mortality in this group was 12.5%. Eight patients received left ventricular aneurysmectomy in addition to valve surgery, three of them died early.We conclude that patients with highly impaired left ventricular function and ischemic mitral insufficiency are at too great a risk for either valve reconstruction or replacement. Cardiac transplantation should be considered for this patient group. However, patients with ischemic mitral insufficiency and moderately impaired left ventricular function can undergo valve reconstruction or replacement with an acceptable prognosis. The goal of mitral valve reconstruction should be reducing mitral valve insufficiency to at least grade I. If this is not achieved, the prognosis after repair is worse than after valve replacement, therefore, the surgeon should replace the valve without delay.