Many clinical trials or population studies have used change in Minnesota Q code, ST-segment depression code or T-wave inversion code as evidence of new myocardial infarction or new coronary heart disease event. Direct electrocardiogram (ECG) waveform comparison is a new standardized procedure for diagnosing interim myocardial infarction from ECGs classified according to the Minnesota code (serial Q-wave pattern change). This procedure was investigated for its application in epidemiologic studies. Use of this procedure in the Multiple Risk Factor Intervention Trial resulted in a 50% increase in the positive predictive accuracy, improved agreement with clinically defined myocardial infarction and a strong independent prognostic association with total and coronary heart disease mortality. Among those with major Minnesota Q-code findings, there was substantial variation in mortality. The 5-year coronary heart disease death rates estimated by life table analysis were 8.5% for those with major serial Q-wave pattern change, 5.1% for those with minor serial Q-wave pattern change and 1.5 to 2.6% for those with major or minor Minnesota Q-code change not substantiated by direct waveform comparison, compared with 2.4% for those with no Minnesota Q-code findings. The coronary heart disease death rate for those with major serial Q-wave pattern change was greater than that for the other ECG groups (p less than 0.01). Adjustment for age and other risk factors did not qualitatively alter these findings. This new approach is eminently suitable for export to other investigators, for incorporation into computer analysis programs and for statistical analysis.