In patients with suspected AMI. Monitoring of a combination of myoglobin and CK-MB or tn-T allowed ruling-in AMI within 2-3 hours and ruling-out AMI within 3-6 hours in almost all patients admitted with chest pain and a nondiagnostic ECG. This might have a large impact on the early handling and treatment of these patients. The neural network methodology, with monitoring of myoglobin, CK-MB and tn-T allowed, within the first three hours, reliable diagnosis/exclusion of AMI/MMD and prediction of infarct size in patients admitted with suspicion of AMI. The computer system was faster than clinicians. Thus, neural network methodology might be a useful support for the early assessment of patients with suspected myocardial infarction. In patients with unstable CAD. The risk of subsequent cardiac events is increased by increasing maximal levels of tn-T obtained during the initial 24 hours. Thereby a normal, a slightly elevated and a clearly elevated tn-T level identified a low, intermediate and high risk group, respectively, for MI or death. The tn-T level was an independent prognostic variable for MI or death in a multivariate analysis comparing other early available risk indicators. Furthermore, tn-T seemed to be superior to CK-MB (mass) for risk stratification. In patients able to perform a predischarge ET both the tn-T level and the ET response were independent prognostic indicators for MI or death. The combination of tn-T and the ET response allowed a further improved risk stratification. In patients with tn-T elevation at inclusion, prolonged dalteparin treatment was beneficial. However, in patients without tn-T elevation, long term dalteparin treatment had no protective effect. Thus, tn-T determination provides independent and important prognostic information in unstable CAD. In the selection of treatment strategy for the individual patient, this simple, inexpensive and early available biochemical test might be useful.