Objective: To investigate the effects of using parecoxib sodium for analgesia on post-hepatectomy cognitive function in elderly patients. Methods: Eighty patients over 60 years of age undergoing elective hepatectomy for liver cancer were randomly assigned into observational group and control group, with 40 cases in each group. Patients in observational group received parecoxib sodium administration after surgery (40 mg, intravenous injection, and once per 12 h for 3 d) and those in control group were given normal saline as placebo, and additional tramadol was used in either group according to need for analgesia to maintain the pain visual analog scale (VAS) not more than a score of 3. The Wechsler Adult Intelligence Scale (WAIS) and Wechsler Memory Scale (WMS) were used to test the incidence of postoperative cognitive dysfunction (POCD) in both groups of patients, and the blood levels of S-100β protein, inflammatory cytokines, cortisol, and ammonia were detected before surgery and on several time points after surgery. Results: The perioperative data between the two groups of patients showed no statistical difference (all P>0.05). The incidence of POCD was significantly lower in observational group than that in control group (7.5% vs. 25%, P<0.05). The levels of cortisol, IL-4 and ammonia had no significant alteration in both group at any postoperative time points compared with their preoperative levels (all P>0.05); the levels of S-100β protein, IL-6, TNF-α, IL-1β and CRP in both groups were significantly increased in different time spans with varying degrees in both groups compared with their preoperative levels (all P<0.05), but the increasing degrees of S-100β protein and IL-6 in observational group were less than those in control group within postoperative 24 and 48 h, respectively (all P<0.05). Conclusion: Parecoxib sodium administration can help improve the post-hepatectomy cognitive function in elderly patients, and the mechanism may probably associated with its decreasing the S-100β and IL-6 levels.