The initial ECGs in 440 patients admitted for suspected acute myocardial infarction were retrospectively analyzed to determine predictive values of these ECGs for acute myocardial infarction and to determine differences in the initial ECG for Q wave and non-Q wave myocardial infarction. One hundred (23%) of the study patients were diagnosed as having an acute myocardial infarction. Acute injury was seen in 47% of these patients (positive predictive value [PPV], 84%; 95% confidence interval [CI], 72% to 92%), ischemia in 15% (PPV, 39%; 95% CI, 24% to 57%), and left ventricular hypertrophy with strain in 11% (PPV, 19%; 95% CI, 4% to 29%). Forty-three patients were diagnosed as having a Q wave infarction and 50 patients as having a non-Q wave infarction. Seventy-two percent of the patients with a Q wave infarction had acute injury as the initial ECG interpretation compared with 38% in the non-Q wave infarction group (P less than .001). In contrast, only 17% of patients with Q wave infarction had an initial ECG interpretation of ischemia or strain as compared with 36% of patients with non-Q wave infarction (P = .03). Because of the relatively high incidence of acute myocardial infarction in patients admitted with an initial ECG interpretation of ischemia or left ventricular hypertrophy with strain, prospective studies must be performed to determine if selective patients with acute ST segment depression or ischemic T wave inversion in the setting of suspected acute myocardial infarction may benefit from early thrombolytic therapy.