Background Dexamethasone has been used in surgical patients to decrease nausea, vomiting, and postoperative pain. However, it is not well studied how much dexamethasone complicates glucose control in diabetic patients and whether this leads to poor surgical outcomes. Methods We analyzed 256 diabetic patients who underwent elective hip and knee arthroplasty and evaluated the groups that received dexamethasone intraoperatively (201 patients), those who received dexamethasone postoperatively (237 patients), and those who did not receive the steroid intraoperatively (55 patients) and postoperatively (19 patients). Results 256 diabetic patients were included in the study. The mean age of the group was 68.7 (SD ± 9-10) years. Patients were divided into 123 males (48%) and 133 females (52%). 174 (78%) patients had a total knee replacement operation, and 82 (32%) patients had total hip replacement operation. The mean hemoglobin A1c was 6.728 (SD ± 0.99). The mean ASA score was 2.86 (SD ± 0.38). 201 (78.5%) patients received preoperative or intraoperative dexamethasone, and 237 (92.6%) patients received it postoperatively. The mean blood glucose for all patients raised from 131.9 to 172.2 mg/dL ( P = .012) postoperatively, 206.1 mg/dL in the first 24 hours, and 146.2 mg/dL ( P = .39) in the second postoperative day. The change was significant in patients who had poorly controlled diabetes ( P < .01) preoperatively. There was no significant difference in our study regarding dexamethasone use and effect on postoperative nausea ( P = 1.0) and vomiting ( P = .52). There was an improvement in pain scores in the patients who received dexamethasone postoperatively which was statistically significant ( P = .054). Conclusion Dexamethasone use in diabetic patients for control of postoperative nausea and vomiting in those undergoing elective total knee and hip arthroplasty had a negative impact on glycemic control specifically in those with poorly controlled diabetes and should be avoided.