<p>The aim of this study was to investigate additional risk stratification benefits of myocardial radiomics derived from coronary CTA (CCTA) over clinical factors, anatomic and hemodynamic CTA findings in patients with stable angina. 2171 multicenter patients with stable angina who underwent CCTA were retrospectively included. The endpoints were major adverse cardiovascular events (MACE). Clinical factors and CTA findings were analyzed to identify the independent predictors. The left ventricular myocardium was segmented from CCTA, and the radiomics features were assessed to build a radiomics signature (RS). A radiomics nomogram (RN) was constructed by combining the independent clinical and CTA predictors with the RS. Previous myocardial infarction, hyperlipidemia, and the change in CT derived fractional flow reserve (ΔFFR_CT) were proven as independent predictors for MACE (C-index, 0.733, test cohort). The RS was significantly associated with MACE (C-index, 0.713, test cohort). By adding the RS to the clinical and CTA predictors, the RN provided a higher C-index (0.788, test cohort), and higher time-AUCs. In the test cohort, the IDIs of RN to the clinical and CTA predictors were 3.10% (95% CI, 0.50–5.60%, <i>P</i> = 0.004), and 5.40% (95% CI, 0.40–10.30%, <i>P</i> = 0.040) for 1-year and 3-year MACE, respectively. In subgroup analyses in patients with diabetes mellitus and hypertension, the RS had additional predictive value for MACE over the clinical and CTA findings. Compared to clinical factors, anatomic and hemodynamic CTA findings, CCTA-derived myocardial radiomics provides incremental prognostic value for the prediction of MACE in patients with stable angina.</p> Graphical abstract <p></p>