Aims: The aims of the study were to investigate the effects of the GLP-1 receptor agonist liraglutide as add-on to metformin on insulin sensitivity (Si) and glucose effectiveness (Sg) in addition to its positive effects on beta-cell function in overweight/obese patients with coronary artery disease (CAD) and type 2 diabetes mellitus (T2DM). Methods: The design of the study was a randomized, double-blind, placebo-controlled, cross-over trial in patients with stable CAD and newly diagnosed well-controlled T2DM. Patients were treated with liraglutide/metformin vs placebo/metformin for a 12 + 12-week period with ≥2-week wash-out. First phase insulin secretion (AIRg), Si and Sg were estimated by the Bergman Minimal Model, enabling calculation of beta-cell function; Disposition Index (DI) = AIRg × Si. A total of 30 patients from among 41 randomized were available for paired analysis. Results: Baseline characteristics were: HbA1c 47 mmol/mol (SD 6), BMI 31.6 kg/m² (SD 4.8), fasting plasma-glucose 6.9 mmol/L (IQR 6.1; 7.4) and HOMA-IR 4.9 (IQR 3.0; 7.5). Liraglutide treatment improved AIRg by 3-fold, 497 mU × L⁻¹ × min (IQR 342; 626, P <.0001) and DI by 1-fold, 766 (SD 824, P <.0001). Despite a significant weight loss of −2.7 kg (−6.7; −0.6) during liraglutide treatment, we found no improvement in HOMA-IR, Si or Sg. Weight loss during liraglutide therapy did not result in a carry-over effect. Conclusion: Liraglutide as add-on to metformin induces a clinically significant improvement in beta-cell function in overweight/obese, high cardiovascular risk patients with newly diagnosed well-controlled T2DM and CAD. The effect of liraglutide on DI is mediated entirely by improved AIRg whereas the effects on Si and Sg are neutral.