This study was performed to elucidate the localization at the cellular level of technetium-99m phosphorus ((99m)Tc-P) radiopharmaceuticals in acute myocardial infarcts and the mechanisms responsible for (99m)Tc-P uptake in acute myocardial infarcts and other tissues. In 20 dogs with proximal left anterior descending coronary arterial ligation for 1-3 days, elevated calcium levels were measured at all sites of increased (99m)Tc-P uptake (acute myocardial infarcts, necrotic thoracotomy muscle, lactating breast, and normal bone); however, a consistent linear relationship between (99m)Tc-P and calcium levels was not observed. A strong correlation (r = 0.95 and 0.99, n = 2 dogs) was demonstrated between levels of (3)H-diphosphonate and (99m)Tc-P in infarcted myocardium. Autoradiographic studies with (3)H-diphosphonate revealed extensive labeling in the infarct periphery which contained necrotic muscle cells with features of severe calcium overloading, including widespread hypercontraction as well as more selective formation of mitochondrial calcific deposits. Autoradiography also demonstrated labeling of a small population of damaged border zone muscle cells which exhibited prominent accumulation of lipid droplets and focal, early mitochondrial calcification. Cell fractionation studies revealed major localization of both (99m)Tc-P and calcium in the soluble supernate and membrane-debris fractions of infarcted myocardium and less than 2% of total (99m)Tc-P and calcium in the mitochondrial fractions; however, electron microscopic examination showed that mitochondria with calcific deposits were not preserved in the mitochondrial fractions. In vitro studies evaluating the role of serum protein binding on tissue uptake of (99m)Tc-P agents demonstrated that, in spite of significant complexing with serum proteins, serum (99m)Tc-P activity retained the ability to adsorp to calcium hydroxyapatite and amorphous calcium phosphate. In vivo studies showed that concentration of human serum albumin (labeled with iodine-131) in infarcted myocardium reached a maximum of only 3.8 times normal after a circulation time of 96 h, whereas (99m)Tc-P uptake was at least 10 times normal after a circulation time as short as 1 h. It is concluded that: (a) (99m)Tc-P uptake in acutely infarcted myocardium, and possibly other types of soft tissue damage, is limited to necrotic and severely injured cells; (b) concentration of (99m)Tc-P results from selective adsorption of (99m)Tc-P with various forms of tissue calcium stores, including amorphous calcium phosphate, crystalline hydroxyapatite, and calcium complexed with myofibrils and other macromolecules, possibly supplemented by calcium-independent complexing with organic macromolecules; and (c) lack of a linear relationship between (99m)Tc-P and tissue calcium levels mainly results from local differences in composition and physicochemical properties of tissue calcium stores and from local variations in levels of blood flow for delivery of (99m)Tc-P agents.