Recently, an association between minimally elevated cardiac troponin levels and cardiovascular risk in the general population has been reported. However, the prevalence and clinical importance of elevated cardiac troponin T (cTnT) levels remain unclear in patients with histories of myocardial infarction (MI). In this study, 1,807 consecutive patients with ST-segment elevation MIs were prospectively studied (77.1% men; mean age 64.4 years). Venous blood samples were obtained in the chronic stage of MI (28 +/- 7 days after onset), and serum cTnT levels were determined. During the average follow-up of 1,042 days, 84 patients died and 83 had nonfatal reinfarctions. Patients with cTnT levels in the highest quartile (> or = 0.040 ng/ml [n = 353]) had a higher incidence of all-cause death (8.2% vs 5.2%, p = 0.049) and nonfatal reinfarction (8.3% vs 5.1%, p = 0.048) than patients with cTnT levels from the lower 3 quartiles (<0.040 ng/ml [n = 1,064]). Multivariate Cox regression analysis revealed that a minimally elevated cTnT level (> or =0.040 ng/ml) was a significant predictor of all-cause mortality (hazard ratio 1.79, 95% confidence interval 1.10 to 2.90, p <0.02) and nonfatal reinfarction (hazard ratio 1.50, 95% confidence interval 1.13 to 2.20, p <0.03). Subgroup analysis showed that an elevated cTnT level was also a predictor of all-cause mortality and nonfatal reinfarction in patients without heart failure. In conclusion, minimally elevated cTnT levels in the chronic stage of MI predicted long-term adverse clinical outcomes.