Although "nontransmural" and "transmural" are morphologic terms used widely to distinguish patients with myocardial infarction, controversy exists as to their meaning regarding clinical course. For this study, a transmural infarct was defined as one that involves essentially the full thickness of the ventricular wall, and nontransmural was defined as something less. The purpose of this study was to identify true morphologic nontransmural acute (less than 21 days old) infarcts at autopsy and compare them with transmural (full-thickness) infarcts in age-matched subjects, for clinical and pathologic similarities and differences. Among the autopsy subjects, comparing 35 nontransmural and 35 transmural infarcts, there was no significant difference with regard to subjects' race or sex, chest pain, arrhythmias, heart block, or cause of death; transmural myocardial infarctions did have a higher frequency of new Q waves (30 of 35 versus six of 35, p less than 0.001) and presented more often with increasing dyspnea. At autopsy, there were no significant differences regarding heart weight, location of infarcts, severity of coronary disease, age of acute infarct, or total size of infarct (18 percent of left ventricle for nontransmural versus 22 percent for transmural). There was, however, a significantly greater tendency for those with nontransmural infarct to have evidence of prior infarction at autopsy (27 of 35 versus 19 of 35, p less than 0.05). Acute coronary thrombi in the distribution of the infarct were significantly more common among transmural myocardial infarcts (32 of 35 versus 18 of 35, p less than 0.001). Morphologically, the nontransmural infarcts showed mural involvement ranging from 20 to 90 percent of the left ventricle, and histologically showed more contraction band (i.e., reflow) injury (57 percent with more than 30 percent contraction band necrosis) compared with transmural infarcts (32 percent with more than 30 percent contraction band necrosis) (p less than 0.05). Fatal nontransmural and transmural infarcts have major clinical and pathologic similarities, but differences in number of prior infarcts, type of necrosis, and occurrence of coronary thrombi suggest differing pathophysiology. The heterogeneity of both transmural and nontransmural infarcts likely accounts for existing differences among clinical studies regarding prognosis. Although this classification system has value in the clinical setting, that at times it represents an imprecise oversimplification of infarct type should be recognized in assessing individual patients.