This study evaluated the incidence and type of acute myocardial infarction (AMI) in a consecutive population with increased troponin I (TnI). AMI has recently been redefined and subclassified. Incidence, demographic data, angiographic findings, and hospital mortality of patients with various AMI subtypes or an increased TnI in the absence of AMI have not been previously reported in a prospective study. Over a 3-month period, all patients admitted from an emergency room or from in-patient services with >1 TnI level >0.04 ng/ml were evaluated and subclassified in AMI subgroups. In-hospital or recent coronary angiograms were reviewed. In-hospital mortality was noted. Of 2,944 patients with serial TnI measurements, 728 had an increased TnI and 701 (23.8%) were evaluated. Two hundred sixteen (30.8% with increased TnI and 42.7% with "rule-out MI" on admission) met criteria for AMI. One hundred forty-three (20.4%) had type 1, 64 (9.1%) had type 2, whereas 461 (65.8%) did not meet criteria for AMI. On multivariate analysis, use of angiography, peak TnI level, hyperlipidemia, and illicit drug use were independently associated with the diagnosis of AMI. TnI of 0.28 ng/ml had a 70% sensitivity and specificity for AMI diagnosis. In conclusion, a minority admitted with increased TnI have AMI by the universal definition. Type 1 is the most common AMI and is associated with higher TnI values and these patients are more likely to undergo angiography. Type 2 AMI is often associated with illicit drug use.