BACKGROUND High-sensitivity cardiac troponin (hs-cTn) may allow an earlier diagnosis of acute myocardial infarction (AMI). METHODS We prospectively enrolled 1148 (derivation cohort) and 517 (external validation cohort) unselected patients presenting with suspected AMI to the emergency department. Final diagnosis was adjudicated by 2 independent cardiologists. Hs-cTnT was measured at presentation and after 2 hours. A diagnostic algorithm incorporating hs-cTnT values at presentation and absolute changes within the first 2 hours was derived. RESULTS AMI was the final diagnosis in 16% of patients in the derivation and 9.1% in the validation cohort. The 2-hour algorithm developed in the derivation cohort classified 60% of patients as "rule-out," 16% as "rule-in," and 24% in the "observational-zone." Resulting sensitivity and negative predictive value (NPV) were 99.5% and 99.9%, respectively, for rule-out, and specificity and positive predictive value (PPV) were 96% and 78%, respectively, for rule-in. Applying the 2-hour triage algorithm in the external validation cohort, 78% of patients could be classified as "rule-out," 8% as "rule-in," and 14% in the "observational-zone." Resulting sensitivity and NPV were 96% and 99.5%, respectively, for rule-out, and specificity and PPV were 99% and 85%, respectively, for rule-in. Cumulative 30-day survival rates were 100%, 98.9%, and 95.2% (P < .001), and 100%, 100%, and 95% (P < .001) in patients classified as "rule-out," "observational-zone," and "rule-in" in the 2 cohorts, respectively. CONCLUSIONS A simple algorithm incorporating hs-cTnT baseline values and absolute changes over 2 hours allowed a triage toward safe rule-out, or accurate rule-in, of AMI in the vast majority of patients, with only 20% requiring more prolonged monitoring and serial blood sampling.