SINCE Abrams (1) first reported the differential response of neoplastic renal vessels to epinephrine in man, there has been growing clinical interest in the direct intra-arterial injection of various vasoactive agents as a diagnostic aid in tumor angiography. This technic is based upon the premise that tumor vessels respond differently from normal vessels to specific pharmacologic agents. Although there is some conflicting experimental evidence on this point (4, 8, 13, 16) the clinical application has been amply demonstrated, particularly in renal tumors (9, 11, 12). Recently attention has been called to the angiographic similarities between tumors and certain benign inflammations (abscess 3, 5, 18; pyomyositis, 6; and osteomyelitis 15) and reparative, processes (organizing hematoma 19; and myositis ossificans 14). Abnormal vessels, capillary stains, and early venous filling have all been observed in these non-neoplastic lesions, and in the kidney the arteriographic differential diagnosis between a poorly vascularized hypernephroma and a renal carbuncle may be very difficult (3, 5). Were it possible to distinguish neoplastic from inflammatory vasculature with vasopressor arteriography preoperatively, the surgical approach to renal masses could be more logically planned. Very little is known, however, about the responsiveness of inflammatory vasculature to pharmacologic agents. Kahn and Wise (10) have recently reported the simulation of a renal tumor response to epinephrine by a chronic renal abscess, leading to an erroneous preoperative arteriographic diagnosis of hypernephroma. For these reasons we have undertaken a laboratory study of the arteriographic patterns in healing wounds and the response of this non-neoplastic vascular bed to intra-arterial vasoconstrictors. Materials and Methods Twelve monkeys (Macaca mulatta) weighing between 3 and 6 kilograms were studied. Using sterile technic, transverse incisions were made in the lower dorsalupper lumbar region of the back and dissection was carried down to the level of the spinal laminae. One lamina was removed for exposure of the cord, and in most animals a costotransversectomy was also performed. The incisions were then closed with conventional surgical technic and allowed to heal normally. There was no deliberate attempt to produce contaminated wounds and in no instance did suppuration or dehiscence occur. Six to ten days postoperatively, arteriographic studies were performed. Under pentobarbital anesthesia, polyethylene catheters (PE 160) were introduced by femoral artery cutdown and positioned in the midthoracic aorta. Care was taken to insure that the catheter tip was well above the level of the wound.