Sensory nerve dysfunction is a common symptom in hemodialysis (HD) patients and to have a more objective diagnose is more and more important for clinician recently. The current perception threshold (CPT) is a method for evaluation sensory nerve function via pain feeling and which has been reported better than the conventional methods. In this study, we proposed a system based on Flexible Fuzzy Classifier (FFC) using CPT index to evaluate sensory nerve dysfunction in HD patients. All patients were recruited at Yong-Kang Veterans Hospital including 35 HD patients (20 male/15 female) versus 20 control patients (10 male/10 female) and measured from arm and above the ankles before HD therapy. The results indicated that CPT levels in HD patients were significantly higher than normal aged subjects (control group), and over 7 years of HD patients were appeared to be higher but not significant with other patients. Furthermore, FFC method could classify HD patients and controls group successfully with 100% accuracy; 82% accuracy for over 7 years HD patients and others. The FFC method with CPT index has high potential to be an automatic classification system for early stage detection of sensory nerve dysfunction.