Considerable emphasis has been placed on the effect of contrast media in increasing the permeability of cerebral vessels and “the blood-brain barrier,” particularly in Broman and Olsson's classical reports of 1948–56 (1–4). The damaging effect of contrast agent is not limited to this area alone, as is shown by the electrocardiographic changes noted during angiocardiography (5, 6) and kidney damage following renal angiography (7, 8). We are all familiar with the increase in blood flow that occurs in peripheral angiography. It is probably produced by the inhibition of smooth muscle tone in vessel walls, and this in turn leads to vasodilatation. Such a response indicates the probability that there is loss of contrast material from the vessel lumen into surrounding tissues. An occurrence such as this has already been recorded by Oppenheimer et al. (9), who demonstrated a persistent contrast blush in the myocardium of dogs following coronary angiography. In this communication we report an attempt to show the leakage of contrast material from the intravascular compartment as a result of increased permeability. Experimental Methods and Results In our experiments the microvessels were observed directly. The extension of the wing of a common brown bat (Myotis lucifugus) over a glass plate permits microscopic observation of the circulation in a living animal undisturbed by surgical procedures and anesthesia. The preparation is comparatively simple in that the extreme thinness of the wing (approximately 17 microns) allows transillumination and visualization of vascular networks with ordinary microscopic equipment. Vessels can be cannulated with a fine glass microcannula utilizing a low-power dissecting microscope. Wiedeman (10) employed this procedure to demonstrate changes brought about by a brief arterial injection of a group of tri-iodobenzoates. Sodium acetrizoate (Urokon) was shown to crenate red blood cells, to produce aggregations of red cells and platelets, to damage the endothelial lining of vessels as evidenced by the adherence of leukocytes to vessel walls, and to cause an initial vasodilatation followed in ten to fifteen minutes by intense vasoconstriction of some of the vessels whose walls had been in direct contact with the bolus of injected material. Other tri-iodobenzoates such as Renografin 60, Renografin 76, and Hypaque 50 resulted in similar changes to a lesser degree. The contrasts producing the least change were Renovist and Hypaque M 90 which caused only vasodilatation and no cellular effects. Two methods were used to demonstrate increased endothelial permeability produced by exposure to a selection of the more popular contrast media. In the first, the agent was introduced intravenously into a side branch from the midportion of the main wing vein, so as to expose only the proximal part of this vein to the material. Injection pressure was controlled visually so that the existing venous pressure was not exceeded.