BACKGROUND Diabetic patients with acute myocardial infarction (MI) have higher mortality than nondiabetic patients. The purpose of this study was to examine if larger infarct size explains the higher mortality in diabetic patients with acute ST-segment-elevation MI. METHODS In the CORE trial (n = 2948), subsets of patients underwent quantitative radionuclide measurement of technetium Tc 99m sestamibi infarct size (n = 1164) or gated equilibrium left ventricular ejection fraction (LVEF) (n = 1137) at days 6 to 16 after thrombolytic therapy. Clinical follow-up was 96.7% complete at 6 months. RESULTS The prevalence of diabetes in these patient imaging subsets was 16% to 17%. Higher risk clinical characteristics including older age and a greater prevalence of prior MI were more common in diabetic patients. Median infarct size was larger in diabetic patients (22% vs 17% of the left ventricle, P = .04), a difference that remained significant after adjustment for clinical variables (P = .048). Patients with diabetes also had lower median LVEF (48% vs 51%, unadjusted P = .002, adjusted P = .007). Six-month mortality was higher in diabetic patients: infarct size subset, 5.9% vs 1.6% (P = .0016); LVEF subset, 6.1% vs 1.0% (P < .0001). Multivariable models demonstrated that diabetes and each imaging variable were independent predictors of mortality. CONCLUSIONS Infarct size is modestly larger and LVEF modestly lower in diabetic patients with ST-segment-elevation MI. The substantially higher (4- to 6-fold) mortality rate in diabetic vs nondiabetic patients is only partially explained by relatively small differences in infarct size and LVEF.