BACKGROUND The World Health Organization (WHO) criteria for the diagnosis of acute myocardial infarction (AMI) includes presentation of chest pain over 20 min, evolutionary changes on the electrocardiogram (ECG), and abnormal levels of cardiac enzymes. OBJECTIVE A multicenter study was conducted to evaluate the efficacy of cardiac troponin I (cTnI) in detecting and ruling out AMI. METHODS The normal range for cTnI in 149 apparently healthy subjects without known history of cardiac or other diseases was 0 to 0.5 ng/ml. Cutoffs of 2.5 ng/ml for c TnI and 5.0 ng/ml for creatine kinase-MB (CK-MB) were used. RESULTS The diagnostic sensitivity of blood collected from 291 consecutive patients with suspicion of AMI was 95.0 and 96.4%, respectively, for samples obtained at 4-48 h after AMI onset. CK-MB was more sensitive during the early 4-8 h interval (84 vs. 74%); both had 100% sensitivity from 12-36 h. CTnI remained at 100% for 72 h, while CK-MB declined to 57%. The clinical specificity was 97.4 vs. 85.8%, respectively, on non-AMI patients with cardiac and noncardiac diseases, and those with renal disease. CONCLUSIONS cTnI is an excellent marker for detecting and ruling out AMI, because it has better specificity and a wider diagnostic window than the accepted standard, CK-MB.