<b><i>Aims:</i></b> This study assessed the validity of the Estimation of Physiologic Ability and Surgical Stress (E-PASS) score in maintenance hemodialysis patients undergoing elective abdominal surgery. <b><i>Methods:</i></b> We retrospectively reviewed the medical records of 73 hemodialysis patients who underwent elective gastrointestinal surgery. The main outcomes analyzed were the E-PASS score and postoperative course, which were defined by mortality and morbidity. The discriminative capability of the E-PASS score was evaluated using receiver operating characteristic (ROC) curve analysis. <b><i>Results:</i></b> The overall mortality rate observed was 2.7% (2 patients) and the morbidity rate was 36.9%. There were no significant differences in the comprehensive risk score, preoperative score or surgical stress score for patients with or without complications (p = 0.556, 0.639 and 0.168, respectively). Subsequent ROC curve analysis demonstrated poor predictive accuracy for morbidity. When the results in our study population were compared with those in Haga's study population, our population exhibited a highly significant rightward shift (p < 0.001). <b><i>Conclusion:</i></b> The E-PASS score was a poor predictor of complications because maintenance hemodialysis patients already have relatively high risk factors. This scoring system should not be applied in such a special group with high risk factors. i 2014 S. Karger AG, Basel