In most theoretical analyses of the heart, the tissue has been assumed to be incompressible. Because the myocardium is extensively perfused with distensible vessels, increasing the stiffness of the surrounding tissue, as with contraction or passive stretching, should decrease the volume of fluid in these vessels. Using a digital subtraction angiographic method, we quantified the amount of vascular volume extruded from six passive, perfused canine interventricular septa during cyclic biaxial loading from 300 to 900 g force. At pressures from 0 to 120 mmHg the amount of fluid extruded during a loading cycle varied from 2 to 4 ml/100 g tissue at 0 and 120 mmHg, respectively. This volume change increased with perfusion pressure and was significantly greater at 120 than at 0, 30, or 60 mmHg. The amount of fluid extruded was on the same order as that estimated during active contraction or with a 60-mmHg change in perfusion pressure. The finding that perfused myocardium is compressible implies that results from existing analyses of the heart assuming incompressibility are not realistic. Such analyses must account for compressibility with, e.g., mixture theory or other similar approaches.