Although glycated hemoglobin A1C (HbA1c) has classically been used for glycemic control screening before surgery, fructosamine, a short-term glucose variability indicator, has been reported to be a more accurate predictor of postoperative periprosthetic joint infection among patients with diabetes mellitus (DM). Given the variability of diabetic medication management, this study aims to identify the associated effect of glycemic control medication regimen (GCMR) on the incidence rate and associated odds of abnormal preoperative fructosamine levels among diabetic primary total knee arthroplasty or total hip arthroplasty patients. Between 2017 and 2018, consecutive series of total hip arthroplasty and total knee arthroplasty patients were identified, and the final cohort included only diabetic patients. All patients reported preoperative HbA1c and fructosamine levels. GCMR categories included insulin, metformin, and other. Independent risk of GCMR and abnormal fructosamine levels (>293 µmol/L) were identified using multivariable logistic regression, while controlling for preoperative baseline factors including HbA1C. Among 420 patients, 15.7% (66/420) were diabetic, of whom 22.7% (15/66) reported an abnormal fructosamine level. Among patients requiring GCMR, 24.0% (18/75), 56.0% (42/75), and 77.7% (58/75) reported using insulin, other, and metformin, respectively. Multivariable logistic regressions demonstrated that insulin-dependent patients with DM reported a 1.71 (95% confidence interval [CI], 0.096 to 30.213, P = 0.716) increased odds of abnormal fructosamine levels compared with nonactive GCMR patients, whereas patients managed with metformin and other glycemic control medications reported a protective 0.48 (95% CI, 0.418 to 5.407, P = 0.549) and 0.32 (95% CI, 0.216 to 4.508, P = 0.393) decreased odds of abnormal fructosamine levels, respectively. In this study, insulin and other GCMR medications exhibited a trend for increased and decreased odds of having abnormal preoperative fructosamine levels while controlling for baseline HbA1c level compared with patients with DM without active GCMR. This association may be explained by multifactorial short-term glucose variability in insulin users, indicating the continued need and optimization of short-term glycemic variations instead of HbA1c.