We examined the feasibility of early imaging of myocardial infarcts by intracoronary injection of 131I-labelled cardiac myosin-specific antibody (Fab')2. The left anterior descending coronary artery was occluded for 5 hours by a balloon catheter introduced through the carotid artery in 12 dogs. The catheter was withdrawn and 1 mCi 201Tl was injected intravenously and 500 muCi of 131I antibody were injected into the main left coronary artery. Six of these animals demonstrated evidence of myocardial infarction by ECG and subsequent triphenyl-tetrazolium chloride staining, while the others did not. In each of the infarcted animals, in vivo scintograms one-half hour after injection of isotope showed uptake of 131I in the anteroapical region of the heart corresponding to the region of absent 201Tl uptake. This relationship was confirmed in the excised hearts and in heart slices. In slices, 131I uptake corresponded to regions that did not stain with triphenyltetrazolium chloride. In the six animals that did not show evidence for infarction after coronary occlusion, uptake of 131I was not demonstrated, either in vivo or in excised specimens. In four additional dogs subjected to the same procedure, 125I-labelled (Fab')2 from nonimmune IgG was injected simultaneously into the left main coronary artery with 131I-labelled canine myosin-specific antibody (Fab')2. The ratio of uptake between infarct center and normal tissue was 34.3 +/- 1.5 (mean+/-SEM) for the specific antibody fragment as contrasted to 6.6+/-0.4 for the nonimmune IgG fragment, indicating that intracoronary injection does not favor nonspecific sequestration of protein in regions of infarction. Thus, the intracoronary administration of myosin-specific antibody fragments leads to early and specific one-half hour imaging of myocardial infarcts.