Based on the STS/AATS guidelines of 1996, we compared the long-term results after mitral (Re-MVR) and/or tricuspid valve re-replacement (Re-TVR) in a total 324 patients (Re-MVT 299.Re-MVT+TVR 19, and Re-TVR 6 patients) with those after initial valve replacement in 763 patients (MVR 741, MVR+TVR 6 and TVR 16 patients). The actuarial survival (AS), reoperation-free (RF), thromboembolism-free (TF), and freedom from all valve-related events (EF) rates at the 15th postoperative year were 69.3%, 82.4% 86.6%, and 48.9% after Re-MVR and 87.2%, 92.2%, 83.9%, and 61.4% after initial MVR, respectively. The only significant difference between the two MVR groups occurred in the RF proportion. Similarly, the incidence of valve-related events after Re-MVR+TVR or Re-TVR was the same as that after the initial operation for the respective valve lesions. These long-term results suggest that valve re-replacement for mitral and/or tricuspid valve lesions should be encouraged to the same extent as the initial operation.