Capillary permeability of 51Cr-EDTA in the canine myocardium was determined by applying: (A) the single injection, external registration method, and (B) the local tissue clearance method, to the intact dog heart of open-chest anesthetized dogs. (A) 51Cr-EDTA was administered into the left anterior descending coronary artery as a bolus injection, and the response curve was recorded by external registration. The capillary diffusion capacity (the permeability-surface area product) for 51Cr-EDTA amounted to 32.5 ml/100 g x min at a capillary extraction of 0.40 and a plasma flow of 73.3 ml/100 g x min. The diffusional permeability coefficient of 51Cr-EDTA was calculated to 1.08 x 10(-5) cm/s, which indicates that the permeability of the capillaries in the myocardium for 51Cr-EDTA is similar to that of continuous capillaries in other tissues. (B) 51Cr-EDTA (3-100 microliters) was injected at a depth of 5 mm into the myocardium of the left ventricular free wall and the residue curve was recorded. The capillary extraction, as determined by the tissue clearance method, was calculated to values 6 times smaller than determined with the single injection, residue detection method. This unreasonably low extraction was probably due to methodological errors inherent to the tissue clearance technique.