BACKGROUND: It is unclear if mild or moderate mitral valve regurgitation (MR) should be repaired at the time of coronary artery bypass grafting (CABG). We sought to determine the long-term effect of uncorrected MR, measured by intraoperative transesophageal echocardiography (TEE), in CABG patients. METHODS AND RESULTS: Between May 1999 and September 2003, data were gathered for 3264 consecutive patients who underwent isolated CABG and had MR graded by intraoperative TEE. MR was graded on the following 5 levels: none, trace, mild, moderate, and severe. Patients who had severe MR or who underwent mitral valve surgery were eliminated from the analysis. The remaining patients were combined into the following 3 groups: none or trace, mild, and moderate MR. Preoperative and follow-up data were 99% complete. The median length of follow-up was 3.0 years. Multivariable analysis controlling for important preoperative risk factors was performed to determine predictors of death and death/hospitalization for heart failure. Increasing MR was a risk factor for death [hazard ratio (HR), 1.44; P<0.001] and death/heart failure hospitalization (HR, 1.34; P<0.01). When patients with moderate MR were eliminated from the analysis, mild MR was a risk factor for death (HR, 1.34; P=0.011) and death/hospitalization for heart failure (HR, 1.34; P<0.001). CONCLUSIONS: Even mild MR, identified by intraoperative TEE, predicts worse outcomes after CABG. Revascularization alone did not eliminate the negative long-term effects of mild MR. CABG patients with uncorrected mild or moderate MR are at increased risk for death and heart-failure hospitalization; consideration for surgical repair or more aggressive medical management and follow-up is warranted.