OBJECTIVE: With the global aging population, the incidence of OA is rising annually, and the number of TKA surgeries is rapidly increasing, placing a heavy economic and healthcare burden on society. As one of the key medications in the ERAS protocol, DXM can significantly reduce postoperative pain, suppress nausea and vomiting, and accelerate patient recovery. However, the safety of perioperative DXM use in patients with diabetes remains unclear. This study aims to clarify the safety of perioperative DXM application in diabetic patients. METHODS: This retrospective analysis involved 285 patients with type 2 diabetes and late-stage knee osteoarthritis who underwent unilateral TKA at the Joint Surgery Center of Peking University Third Hospital from January 2019 to November 2022. After application of the inclusion and exclusion criteria, 161 patients were included in the study. The patients were divided into two groups according to whether they had received continuous intravenous administration of DXM for 3 days postoperatively: the DXM group (n = 66) and the non-DXM group (n = 95). All other treatments and medications were the same in both groups. The patients' PBG, incidence of PONV, length of hospital stay, pain scores, and clinical data were collected and compared between the two groups. RESULTS: There were no significant differences in the general preoperative data between the DXM and non-DXM groups. The average PBG and the proportion of patients with levels exceeding 200 mg/dL were not significantly different between the two groups (10.84 mg/dL vs. 11.05 mg/dL and 43.2% vs. 43.9%). The postoperative visual analog scale scores (3.67 vs. 2.48) and the incidence of PONV were significantly lower in the DXM than non-DXM group (40% vs. 16%). The preoperative level of glycated hemoglobin accurately predicted PBG. Furthermore, there were no statistically significant differences in the incidence of early postoperative complications between the groups. CONCLUSIONS: The administration of DXM after unilateral TKA can effectively reduce postoperative pain and suppress the occurrence of PONV without affecting PBG in patients with type 2 diabetes. In addition, the preoperative level of glycated hemoglobin can accurately predict PBG.