A total of 313 consecutive patients was studied to assess the prevalence and prognostic implications of Q-wave loss after transmural acute myocardial infarction. Heart catheterization, including single-plane left ventriculography and selective coronary arteriography, was performed before hospital discharge. After a mean follow-up of 65 (1 to 100) months, 34 patients (11%) lost their Q waves. The time interval from the acute event to the first electrocardiogram showing Q-wave disappearance was 14 (1 to 32) months. Peak creatine kinase value was significantly higher in patients who retained their Q waves than in those who lost them (1,121 +/- 813 vs 779 +/- 464 IU, respectively, p less than 0.05). Severity of coronary artery disease, as judged by the number of diseased arteries and the number of arteries with total or subtotal occlusion, was similar in both groups. However, patients showing Q-wave regression had lower left ventricular end-diastolic pressure, higher ejection fraction and fewer abnormally contracting segments than their counterparts (12 +/- 6 vs 15 +/- 7 mm Hg, p less than 0.05; 53 +/- 11 vs 44 +/- 14%, p less than 0.001; 1 +/- 1 vs 2 +/- 1 segments, p less than 0.001, respectively). In addition, no patient with normalized electrocardiogram presented with left ventricular aneurysm. Although differences in mortality, nonfatal reinfarction and new onset of angina between the 2 groups were not significant, congestive heart failure was prevalent among patients with permanent Q waves (23 vs 6%, p less than 0.05). Our findings suggest that Q-wave loss after AMI may be related to a smaller infarct size.