Objective: To explore the mechanisms of electro-acupuncture (EA) induced sedation during general anesthesia through investigating the role of the benzodiazepines-GABA signaling pathway. Method: 80 patients undergoing abdominal surgery were randomly divided into four groups (n=20): C group (propofol only group), F group (propofol + flumazenil group), EA group (propofol + electro-acupuncture group), and EA + F group (propofol + electro-acupuncture + flumazenil group). Before induction, acupuncture needles were inserted at the points of Zusanli (ST 36) and Neiguan (PC 6) bilaterally to elicit "DeQi". Target-controlled infusion (TCI) of propofol was used for the induction and maintenance of anesthesia. After 15 minutes, equilibrium of Narcotrend Index (NT index) was achieved; patients were then assigned to receive different interventions. In EA group and EA + F group, patients received EA for 30 minutes under general anesthesia, and flumazenil with the dosage of 0.1 mg/kg was added in the latter at the end of EA. In C group and F group, patients did not receive EA, and flumazenil with the dosage of 0.1 mg/kg was administered in F Group at the same time point as EA + F group. Heart rate (HR), mean arterial pressure (MAP) and NT index were recorded at different time points. Result: There was a significant decrease in NT index of EA group 15 minutes after EA stimulation compared with control C group (P<0.05). To our surprise, after administering flumazenil, there was no significant difference in NT index between EA group and EA + F group (P>0.05). The HR and MAP between the two groups also showed no significant difference (P>0.05). Conclusion: The drop of NT index which induced by EA didn't reverse after administration of flumazenil. The effect of EA in deepening anesthesia may not be mediated by the benzodiazepines-GABA signaling pathway.