This report represents the first application of immunohistochemical methods for localizing an exogenously administered drug. Intravenously administered procainamide was localized in normal, ischemic, and necrotic myocardium in 23 dogs. Rabbit antiprocainamide antibodies were used in an avidin-biotin-peroxidase complex staining method. Normal myocardium demonstrated diffusely positive immunostaining for procainamide, as did the cardiac conduction system and vascular endothelial cells. Necrotic myocardium demonstrated markedly reduced to absent immunostaining. By contrast, in regions of myocardial ischemia without necrosis, immunostaining for procainamide was similar to that in the normal myocardium. Procainamide myocardial tissue levels were reduced in necrotic and ischemic zones compared to normal (p less than 0.05) only in those animals in which procainamide was administered after rather than before the onset of coronary occlusion. The demonstration of the absence of drug binding in the necrotic cells suggests that myocardial tissue levels or radiolabelled assessment of drug distribution can be misleading when nonhomogeneous tissue is sampled. The immunohistochemical technique provides additional information about the regional and cellular distribution of procainamide that is complementary to the information obtainable by radiolabelling microspheres and from biochemical assays.