OBJECTIVE: Analyze the risk factors for hospital mortality in valvar reoperations. METHOD: A prospective analysis was performed of 194 patients that underwent valvar reoperations between July 1995 and June 1999. The following variables were analyzed: gender, age, functional class, number and type of previous operations, cardiac rhythm, urgency at operation, creatinine level, left ventricular ejection fraction, left ventricular systolic and diastolic diameters, right ventricular systolic pressure, prothrombin activity, activated partial thromboplastin time relation, platelet count, cardiopulmonary bypass time, aortic cross-clamp time, number and position of valves, type of procedure, associated procedures and intraoperative bleeding volume. Univariate and multivariate statistical analyses were performed to determine the risk factors for hospital mortality. RESULTS: The overall hospital mortality was 8.8% (17 patients). Univariate analysis showed that the following variables were associated with higher mortality rates: advanced New York Heart Association functional class, decreased left ventricular ejection fraction, decreased prothrombin activity, increased creatinine level, longer aortic cross-clamping time, prolonged cardiopulmonary bypass time, concomitant associated procedures, and higher intraoperative bleeding volume. Logistic multivariate analysis identified advanced New York Heart Association functional class, creatinine level higher than 1.5 mg/dl, and cardiopulmonary bypass time longer than 120 minutes as independent predictors of hospital mortality. CONCLUSIONS: The variables functional class IV, creatinine level > 1.5 mg/dl and cardiopulmonary bypass time > 120 min were independent predictors of hospital mortality in valvar reoperations.