A retrospective study of 100 consecutive cases of autopsy-proved acute myocardial infarction has disclosed a surprisingly low frequency (53%) of correct antemortem diagnoses. Incorrect diagnoses seemed to be caused by (1) unjustified dependence on misleading laboratory studies, (2) inattention to suggestive or diagnostic laboratory studies, (3) atypical or obscure presentation of myocardial infarction, and (4) failure to consider acute myocardial infarction as a diagnostic possibility, particularly when the responsible physician was not a specialist in internal medicine or its subspecialty of cardiology. Suggested corrective measures are a more appropriate use of the ECG and the laboratory and education and reeducation of selected groups of physicians.