Background— Persons with prediabetes and diabetes mellitus are at high risk for cardiovascular events. However, the relationships of prediabetes and diabetes mellitus to the development of subclinical myocardial damage are unclear. Methods and Results— We measured cardiac troponin T with a highly sensitive assay (hs-cTnT) at 2 time points, 6 years apart, among 9051 participants of the community-based Atherosclerosis Risk in Communities Study with no diabetes mellitus, or prediabetes, and without cardiovascular disease including silent myocardial infarction by ECG. First, we examined the incidence of elevated hs-cTnT (≥14 ng/L) at 6 years of follow-up. Second, we examined clinical outcomes during the subsequent ≈14 years of follow-up among persons with and without incident elevations in hs-cTnT. Cumulative probabilities of elevated hs-cTnT at 6 years among persons with no diabetes mellitus, prediabetes, and diabetes mellitus were 3.7%, 6.4%, and 10.8%, respectively. Compared with normoglycemic persons, the adjusted relative risks for incident elevated hs-cTnT were 1.40 (95% CI, 1.08–1.80) for prediabetes and 2.47 (95% CI, 1.78–3.43) for diabetes mellitus. Persons with diabetes mellitus and incident elevations in hs-cTnT were at a substantially higher risk of heart failure (hazard ratio, 6.37 [95% CI, 4.27–9.51]), death (hazard ratio, 4.36 [95% CI, 3.14–6.07]), and coronary heart disease (hazard ratio, 3.84 [95% CI, 2.52–5.84]) compared with persons without diabetes mellitus and no incident elevation in hs-cTnT. Conclusions— Prediabetes and diabetes mellitus were independently associated with the development of subclinical myocardial damage, as assessed by hs-cTnT, and those persons with evidence of subclinical damage were at highest risk for clinical events. These results support a possible deleterious effect of hyperglycemia on the myocardium, possibly reflecting a microvascular cause.