Objective: To investigate the association between the muscle-to-fat ratio (MFR) and metabolic dysfunction-associated fatty liver disease (MAFLD) in patients with type 2 diabetes mellitus (T2DM), and to evaluate its gender differences and diagnostic performance. Methods: A retrospective study was conducted on T2DM patients hospitalized in the Department of Endocrinology at the First Affiliated Hospital of Nanjing Medical University from September 2023 to August 2024. Clinical data were collected from the electronic medical record system. Body composition was measured by bioelectrical impedance analysis, and MFR was calculated as total skeletal muscle mass (kg)/total fat mass (kg). We measured controlled attenuation parameter (CAP) and liver stiffness measurement (LSM) via FibroScan. Patients were divided into the non-MAFLD and MAFLD groups a ccording to whether they had MAFLD. MFR was stratified into gender-specific tertiles. Its association with MAFLD was analyzed using logistic regression and trend tests, with diagnostic performance evaluated via receiver operating characteristic (ROC) curve analysis. Results: A total of 2 056 T2DM patients were screened, of whom 448 met the inclusion criteria and were enrolled. The mean age was 55.33 ± 12.00 years, and the mean body mass index (BMI) was 25.13 ± 3.51 kg/m². The overall prevalence of MAFLD was 62.5% (280/448). The prevalence was lower in males (59.9%, 187/312) than in females (68.4%, 93/136). The MFR was significantly lower in the MAFLD group (n=280) than in the non-MAFLD group (n=168)[1.01 (0.77, 1.24) vs. 1.29(0.96, 1.74), P < 0.001]. MFR tertile analysis showed a decreasing prevalence of MAFLD with increasing MFR (P < 0.001). Multivariable logistic regression revealed gender-specific associations. Compared to the highest MFR tertile, males in the lowest and middle tertiles had a significantly higher risk of MAFLD[OR (95% CI): 3.36(1.49-7.57)and 2.03(1.02-4.03), respectively], with a significant linear trend (P for trend = 0.001). In contrast, among females, only the middle tertile showed a significantly increased risk [OR (95% CI): 3.53 (1.05-11.88)], with no linear trend (P for trend = 0.699). Additionally, mediation analysis demonstrated a statistically significant indirect effect of BMI in females (effect=- 3.59, Bootstrap 95% CI: -7.23--1.09). The area under the ROC curve of MFR for diagnosing MAFLD was 0.766 (cut-off value: 1.22) in males and 0.711 (cut-off value: 0.78) in females. Correlation analysis indicated that MFR was negatively correlated with obesity indices, homeostatic model assessment of insulin resistance, CAP, and LSM in both genders. However, age was negatively correlated with MFR only in females, whereas blood lipids, uric acid, and liver enzymes showed inverse correlations only in males. Conclusion: MFR is independently associated with MAFLD risk in patients with T2DM, and this association exhibits distinct gender-specific patterns, highlighting the importance of considering gender in clinical management.