Background: Cardiac fat is a cardiovascular biomarker but its importance in patients with type 2 diabetes is not clear. The aim was to evaluate the predictive potential of epicardial (EAT), pericardial (PAT) and total cardiac (CAT) fat in type 2 diabetes and elucidate sex differences. Methods: EAT and PAT were measured by echocardiography in 1030 patients with type 2 diabetes. Follow-up was performed through national registries. The end-point was the composite of incident cardiovascular disease (CVD) and all-cause mortality. Analyses were unadjusted (model 1), adjusted for age and sex (model 2), plus systolic blood pressure, body mass index (BMI), low-density lipoprotein (LDL), smoking, diabetes duration and glycated hemoglobin (HbA_1c) (model 3). Results: Median follow-up was 4.7 years and 248 patients (191 men vs. 57 women) experienced the composite end-point. Patients with high EAT (> median level) had increased risk of the composite end-point in model 1 [Hazard ratio (HR): 1.46 (1.13; 1.88), p = 0.004], model 2 [HR: 1.31 (1.01; 1.69), p = 0.038], and borderline in model 3 [HR: 1.32 (0.99; 1.77), p = 0.058]. For men, but not women, high EAT was associated with a 41% increased risk of CVD and mortality in model 3 (p = 0.041). Net reclassification index improved when high EAT was added to model 3 (19.6%, p = 0.035). PAT or CAT were not associated with the end-point. Conclusion: High levels of EAT were associated with the composite of incident CVD and mortality in patients with type 2 diabetes, particularly in men, after adjusting for CVD risk factors. EAT modestly improved risk prediction over CVD risk factors.