It has been suggested that a tentative preoperative decision favoring a pericardial lesion, on the one hand, or a myocardial lesion, on the other, may be made from certain noninvasive procedures, including the electrocardiogram. An attempt was therefore made to detect associated myocardial fibrosis by electrocardiogram in 67 patients with constrictive pericarditis as proven at catheterization (63 patients), surgery (64 patients) or postmortem examination (12 patients). Seven of the 67 had electrocardiograms characteristic of, and 16 compatible with, old myocardial infarct. The electrocardiographic experience was otherwise typical of the literature with non-specific changes in the T waves or RS-T segments and/or low voltage in the remaining 44. All three autopsied patients whose electrocardiograms were interpreted as diagnostic of an old myocardial infarct and both autopsied patients with electrocardiograms compatible with that diagnosis showed myocardial fibrosis. In seven autopsied cases with non-specific T waves or low voltage, the myocardium was normal in three while four showed myocardial fibrosis. It appears that in a few cases right ventricular hypertrophy might have simulated infarct by inducing tall R waves over the right precordium, or R waves which decreased in amplitude as the electrode was passed from the right to the left precordium. Pathologic evidence related myocardial fibrosis to: (1) direct subepicardial penetration by the inflammatory process or deposit of fat in the subepicardial myocardium; (2) compromise of coronary blood flow, as by (a) direct throttling of coronary arteries by scar tissue or (b) deficient irrigation of subendocardial layers due to rigidity of the pericardium; or (3) a concomitant myocardial and pericardial process (lupus, radiation fibrosis, rheumatoid). Independent pericarditis and coronary disease was surprisingly rare. This limited experience (1) suggests that, though myocardial fibrosis may be predicted in constrictive pericarditis if the electrocardiogram shows characteristic changes of myocardial infarction, non-specific T wave changes or low voltage may likewise be associated with myocardial fibrosis, and (2) emphasizes that the difficulty in determining the site of a constrictive process may be compounded by the co-existence in the same heart of both a pericardial and a myocardial process.