Nine hundred sixty-four patients undergoing single valve replacement (SVR) procedures alone or in combination with other cardiac surgical procedures were analyzed prospectively to determine whether the risks of operative mortality (OM) are different when the operation is performed by a resident under direct supervision of an attending surgeon (group I, 49.5%) or by an attending surgeon (group II, 50.5%). The overall OM rate was 8.4% (7.2% for group I and 9.7% for group II, which was not significant). Fifty-eight clinical, angiographic, and hemodynamic variables were analyzed to adjust statistically for differences in patient characteristics in each group. For aortic valve replacement, group II patients had smaller valve area and a higher incidence of three-vessel coronary artery disease, and more patients required resection of an ascending aortic aneurysm, whereas group I patients had higher incidence of peripheral edema and S3 gallop. For mitral valve replacement, more patients in group II had pleural effusion, previous cardiac operations, higher mean pulmonary artery pressures, mitral insufficiency due to coronary artery disease, and were in New York Heart Association (NYHA) functional class 3 and 4, whereas the group I patients had higher incidence of S3 gallop and elevated serum creatinine. Expected OM for each group was derived from a multivariate logistic model. Ratio of observed-to-expected (O/E) OM was calculated to adjust observed OM according to the risk of patients. There was no difference in O/E mortality ratio between the two groups. We conclude that OM probability is not increased when the SVR procedure is performed by residents under direct supervision of attending surgeons when adjusted for patient- and disease-related risk factors. These findings also constitute compelling evidence that patient care does not suffer when resident teaching experience is enhanced in teaching institutions.