Abstract: Objective To evaluate the prediction validation of European system for cardiac operative risk evaluation （EuroSCORE） in prolonged intensive care unit (ICU) stay, mortality, and major postoperative complications for Chinese patients operated for acquired heart valve disease. Methods Between January 2004 and January 2006, 2 218 consecutive patients treated for acquired heart valve diseases were enrolled in Fu Wai Hospital. All these patients accepted valvular surgery. Both logistic model and additive model were applied to EuroSCORE to evaluate its ability in predicting mortality, prolonged ICU stay and major postoperative complications of patients who had undergone heart valve surgery. An receiver operating characteristic curve（ ROC) area was used to test the discrimination of the models. Calibration was assessed by HosmerLemeshow goodnessoffit statistic. Results Discriminating abilities of logistic and additive EuroSCORE algorithm were 0.710 and 0.690 respectively for mortality, 0.670 and 0.660 for prolonged ICU stay, 0.650 and 0.640 for heart failure, 0.720 and 0.710 for respiratory failure, 0.700 and 0.740 for renal failure, and 0.540 and 0.550 for reexploration for bleeding. There was significant difference between logistic and additive algorithm in predicting renal failure and heart failure (Plt;0.05). Calibration of logistic and additive algorithm in predicting mortality, prolonged ICU stay and major postoperative complications were not satisfactory. However, logistic algorithm could be used to predict postoperative respiratory failure (P=0.120). Conclusion EuroSCORE is not an accurate predictor in predicting mortality, prolonged ICU stay and major postoperative complications, but the logistic model can be used to predict postoperative respiratory failure in Chinese patients operated for acquired heart valve diseases.