An experimental in vivo model of aortic aneurysm was established by perfusing an isolated segment of rat abdominal aorta with pancreatic elastase. Ten rats were used in each protocol. Saline-perfused aortas developed no aneurysmal dilations. Elastase-perfused aortas contained aneurysms in the perfused area and a total loss of elastic tissue. Control aortas contained no elastic tissue lesions. There was a quantitative relation between the amount of elastase perfused and aneurysm formation: 1-2 units induced neither macroscopic nor microscopic lesions; 3-6 units induced microscopic elastic tissue damage without macroscopic aneurysm; and more than 6 units produced aneurysmal dilation in all cases. In situ elastase secretion by macrophages was induced by perfusing rat aortas with thioglycollate-activated macrophages or with thioglycollate alone. There was aortitis without true aneurysm and a total loss of elastic tissue in the vicinity of activated macrophages within the aortic media. Perfusion of infra-aneurysmal amount of elastase (1 or 2 units) or thioglycollate plus plasmin (2 units) always induced a large aneurysm, whereas plasmin alone induced neither macroscopic nor microscopic lesions. These morphological results were supported by the significantly elevated elastolytic activity within the aortic wall of animals perfused with thioglycollate plus plasmin 9 days, after perfusion (207.6 +/- 54.8 micrograms elastin-rhodamine lysed/18 hr; control rats, 25.43 +/- 11.13). The results suggest that the presence of elastase within the aortic media leads to aneurysm formation. Activated macrophages within the aortic media may be responsible for elastase secretion and elastic tissue destruction. Plasmin may enhance elastase activity and aggravate the aneurysmal lesion.