The aim of this study was to investigate the mechanism of dexmedetomidine in improving postoperative cognitive dysfunction from the perspectives of alleviating neuronal mitochondrial membrane oxidative stress and electrophysiological dysfunction. A total of 120 patients undergoing elective surgery under general anesthesia from June 2013 to May, 2016 were selected as the subjects of the study and randomly divided into the propofol + remifentanil and dexmedetomidine groups. The Rey Auditory Verbal Learning Test (AVLT) and Beck Depression Inventory (BDI) were performed at day 1 before operation and at day 1, 3, 5 and 15 after operation. The mitochondrial membrane potential was detected using a flow cytometer after staining and labeling for mitochondria in leukocytes via JC-1 fluorescence staining using a fluorescence probe at day 1 before operation and at day 1, 3, 5 and 15 after operation. The activities of mitochondrial respiratory chain complexes at day 1 before and after operation were detected via enzyme-linked immunosorbent assay (ELISA). The results showed that there were no statistically significant differences in the comparisons of general conditions (age, body weight, sex ratio, body mass index, anesthesia time, operation time, and length of stay in the ICU and hospital) for the dexmedetomidine and propofol + remifentanil groups (P>0.05). At day 3 and 5 after operation, the National Institutes of Health Stroke Scale (NIHSS) scores and AVLT scores in the two groups were decreased in different degrees, but the decrease range in the dexmedetomidine group was smaller than that in the propofol + remifentanil group, and the differences were statistically significant (P<0.05). At day 3, 5 and 15 after operation, the BDI scores of the two groups were increased in different degrees, but the increase range in the dexmedetomidine group was smaller than that in the propofol + remifentanil group, and the differences were statistically significant (P<0.05). At day 1, 3 and 5 after operation, the mitochondrial membrane potentials of the two groups were decreased in different degrees, but the decrease range in the dexmedetomidine group was smaller than that in the propofol + remifentanil group, and the differences were statistically significant (P<0.05). The mitochondrial membrane potentials of the two groups returned to the preoperative levels at day 15 after operation. The activities of mitochondrial respiratory chain complex I–IV in the propofol + remifentanil group at day 1 after operation were significantly decreased compared with those before operation, and the differences were statistically significant (P<0.05). The decrease in activities of mitochondrial respiratory chain complex I–IV in the propofol + remifentanil group at day 1 after operation was more significant than that in the dexmedetomidine group, and the difference was statistically significant (P<0.05). The results suggest that dexmedetomidine can relieve neuronal damage that may be caused by mitochondrial membrane oxidative stress, alleviate the damage to mitochondrial related enzyme system activity, and reduce the damage to the activities of mitochondrial respiratory chain enzyme complex I, II, III and IV, ultimately improving the postoperative cognitive dysfunction of patients.