Some of the mechanical and biological problems surrounding the use of fresh allograft inverted aortic valves as mitral valve substitutes are described. Certain aspects of the problem have been studied expzrimentally. In three sheep 'fresh' aortic valve allografts were -inserted, using cardiopulmonary bypass, into the main pulmonary artery, and were observed from 5 to 7 months after operation. The animals survived normally. Their normal pulmonary valves remained in situ. The technique is described. At subsequent necropsy, macroscopically the valves were found to be free from vegetation, and the cusps were pliable and apparently normal. Microscopically, the supporting allograft myocardium showed necrosis and early calcification. The valve cusp showed hyalinization of collagen, although beneath the endocardium this hyalinized collagen contained moderate numbers of fibroblasts with no evidence of proliferation. The endocardium and arterial intima of the allograft showed evidence of ingrowth from adjacent normal host endocardial tissues. The allograft itself was invested in a loose layer of fibro-fatty tissue, which, in view of the necrotic state of the graft myocardium, could well have been a reparative reaction rather than a homograft reaction. It is concluded that, although the cusps could function normally, the necrosis of the myocardium might in time lead to late failure of the graft. Further studies with the valve inserted at mitral level are indicated.In earlier communications Borrie and Mitchell (1960) described the use of the sheep in surgical research, and later indicated that it was useful for studying mitral allograft (homograft) valves implanted under conditions of total pulmonary bypass (Borrie, Lichter, and Miller, 1967).These studies on the mitral valve performed in 1966 and tabulated in the second paper had shown that an inverted, fresh allograft aortic valve, inserted via a left thoracotomy and atriotomy, with the animal placed on cardiopulmonary bypass and moderate hypothermia, could readily be implanted at the site of an excised mitral valve. Two techniques for implanting had been used. In the first, the coronary artery ostia of the graft were opened to give three long strips of aortic wall and a broad valvular face presenting towards the left ventricle of the host. These long ends of aortic wall were sutured to the left ventricular wall at the site of the divided papillary muscles. The base of the inverted valve was attached by continuous suturing, reinforced by locking interrupted sutures, to the mitral annulus. In the second technique the coronary artery ostia of the graft were closed by purse-string sutures, and the val-ve was attached below to the mitral annulus and above to the left atrial wall.During the attachment of the allograft valve to the base of the host's anterior mitral leaflet the aorta was temporarily cross-clamped, for we had tested experimentally that the sheep heart will regularly tolerate cardiac anoxia from aortic cross-clamping for 30 minutes at 320 C. with subsequent compl...