The study investigated the relationship between short-term glycemic variability and subclinical myocardial injury (SMI) in 324 hospitalized patients with type 2 diabetes in a Chinese hospital in 2021–2023. SMI was defined as an hs-cTnT level above 14 ng/L without signs of ischemia or ECG abnormalities and affected 128 patients (39.5%). Short-term glycemic variability was measured by standard deviation (SD) and coefficient of variation (CV) from capillary glucose measurements during hospitalization. In multivariate analysis, independent predictors of SMI were age (OR=1.05, P=0.001), BMI (OR=1.10, P=0.006), duration of diabetes (OR=1.06, P=0.004), insulin use (OR=1.72, P=0.010), SD glucose (OR=1.015, P<0.05), systolic blood pressure (OR=1.02, P=0.038), hs-CRP (OR=1.08, P=0.009) and eGFR (OR=0.97, P=0.003). The predictive model had AUC=0.832 (95% CI 0.787–0.877) and good calibration (Hosmer–Lemeshow P=0.47); the nomogram enables individualized risk estimation. Short-term glycemic variability (SD glucose) was independently associated with SMI, but the study does not demonstrate a causal relationship.