In mortality studies of patients after acute myocardial infarction (AMI), exclusion of patients during selection from the screened population may be important for evaluating the impact of trials, but data on patients excluded from studies are rarely presented. In the Trandolapril Cardiac Evaluation (TRACE) trial of the angiotensin-converting enzyme inhibitor trandolapril versus placebo in patients with left ventricular (LV) systolic dysfunction shortly after AMI, medical history, infarct complication, and survival were accounted for in all patients screened for entry. A total of 7,001 consecutive enzyme-confirmed AMIs were screened for entry in 27 Danish coronary care units. The 1-year mortality of all screened AMI cases was 23% (95% confidence interval 22% to 24%). The target population of the TRACE trial were patients with LV systolic dysfunction (echocardiographically determined wall motion index < or = 1.2, n = 2,606 within 6 days of AMI. The 1-year mortality of this group was 34 +/- 2%. Patients with wall motion index > 1.2 (n = 3,920) had a 1-year mortality of 12 +/- 1%. were excluded. A total of 1,749 were included in the study. The excluded and included groups had a 1-year mortality of 54 +/- 3% and 24 +/- 2%, respectively. The result of the TRACE study will be applicable to two thirds of the patients with LV systolic dysfunction; however, even with this high figure, care should be taken in extrapolating the result to the general population with reduced LV function after AMI since the group excluded from the study had a higher mortality than those who were included.