Background Cardiac troponin T (TnT) is a highly sensitive and specific marker for myocardial damage and can be detected early after myocardial injury. Our hypothesis was to use TnT as an objective marker to verify acute myocardial infarction before hospital admission. Methods and results We evaluated the sensitivity of a rapid qualitative assay for serum TnT for the detection of acute myocardial infarction in the ambulance and assessed the predictive value of a positive prehospital TnT test for death and myocardial infarction during 6-months of follow-up. The study, conducted in an urban area, included 158 consecutive patients with suspected acute myocardial infarction (93 men aged 69 ± 13 years). A myocardial infarction was confirmed in 40 and excluded in 118 patients. The prehospital TnT test was positive in 11 patients, of whom 7 had acute myocardial infarction. Fifty-three patients had a positive test result at hospital admission, with evidence of myocardial infarction in 39 of them. The sensitivity to acute myocardial infarction was 18% for the prehospital and 98% for the in-hospital test with 78% and 88% specificity, respectively. During follow-up, patients with a positive prehospital TnT test result had cardiac events more often (9 of 11) than patients with a negative result (26 of 147; P < .0001). Conclusions In areas with short transport times to the patient the rapid TnT test performed at the point of care identified only a minority of the patients with acute myocardial infarction. A positive prehospital TnT test result seems to be an objective marker for a worse outcome in patients presenting with suspected acute myocardial infarction. (Am Heart J 1999;138:45-8.)