Objective To explore the correlation of muscle mass and strength with diabetes mellitus in community-dwelling middle-aged and older Chinese population. Methods Based on the baseline data of the China Health and Retirement Longitudinal Study (CHARLS) in 2011,8680 eligible subjects aged ≥45 years with complete data were enrolled, including 1461 subjects diagnosed with diabetes mellitus. The general data and laboratory indicators of all participants were collected. Their muscle mass is represented by appendicular skeletal muscle mass (ASM), and hand grip strength (HGS) is measured using a grip meter. The skeletal muscle mass adjusted by body weight (ASM/BM) and hand grip strength adjusted by body mass (HGS/BM) were both categorized into tertiles (Q1, Q2 and Q3), and were assigned 2, 1 and 0 points, respectively. The composite score was calculated as the sum of ASM/BM and HGS/BM scores. The effect of ASM/BM and HGS/BM on risk for diabetes was investigated by logistic regression analysis, and the dose-response relationship of ASM/BM and HGS/BM with diabetes risk was fitted by restricted cubic spline curve. SPSS statistics 22.0, R 4.2 were used for statistical analysis. Data comparison between two groups was performed using student′s t test, Mann-Whitney U test or Chi-square test depending on data type. Results The ASM/BM and HGS/BM were significantly lower in the diabetic patients than the non-diabetic subjects [(27.49±3.90)% vs (29.30±4.51)%, (0.52±0.16) vs (0.57±0.16); P<0.05]. Multivariate logistic regression analysis showed that compared with the individuals from the Q3 group of ASM/BM and HGS/BM, those in Q1 group had 1.260-time (95%CI 1.025~1.548; P<0.05) and 1.496-times (95%CI 1.270-1.763; P<0.05) higher risk of diabetes, respectively. Moreover, the participants in the highest composite score group had a 1.743-times (95%CI 1.333-2.279; P<0.05) higher risk of diabetes than those in the lowest score group. Restricted cubic spline curve revealed that both ASM/BM and HGS/BM had a linear dose-response relationship with the risk of diabetes (non-linear P=0.138,0.570). Subgroup analysis showed that higher composite score significantly increased the risk of diabetes in the non-obese population (P<0.05), but not in the obese population. Conclusion In community-dwelling middle-aged and older Chinese population, muscle mass and strength are influencing factors for diabetes. The higher the composite score is, the higher risk of diabetes will be, especially in the non-obese people.