Because of experience gained in reconstructive mitral valve surgery, we have reevaluated the implantation of cryopreserved homografts in the mitral position. Forty-three patients, aged 11 to 69 years (mean 34 years), underwent mitral valve replacement with cryopreserved mitral homografts. The indications for the procedure were acute endocarditis ( n = 14), rheumatic stenosis ( n = 26), systemic lupus endocarditis ( n = 2), and marasmic endocarditis ( n = 1). All homografts were obtained from hearts explanted in the course of transplantation and were cryopreserved at -160° C in 10% dimethyl sulfoxide solution without antibiotics. Appropriate sizing was based on morphologic study of the homografts and preoperative echocardiographic assessment of the recipient valve. In 82 homografts analyzed, the height of the anterior leaflet was 25 ± 3 mm and the distance from the anulus to the apex of the anterior papillary muscle was 21 ± 3 mm. The morphologic features of the papillary muscles were classified according to four types of increasing complexity. Nine valves with complex (type IV) papillary muscle abnormalities were discarded. Echocardiographic measurements of the valve were matched with those of the homograft identification cards and a slightly larger homograft was selected (measurements + 3 mm). Partial homograft replacement was done in case of a localized lesion (abscess or calcification) ( n = 21). Total homograft replacement was undertaken in the presence of diffuse lesions ( n = 22). Two hospital deaths occurred as a result of poor cardiac output. One patient required reoperation on the tenth postoperative day after a dehiscence on the valvular suture line. After a mean follow-up of 14 months, there has been one late death caused by a bronchial neoplasm and one reoperation for residual stenosis (partial replacement). The remaining patients were in either New York Heart Association class I ( n = 25) or II ( n = 13). Thirty-three patients were in sinus rhythm. Follow-up echocardiography has revealed no mitral regurgitation ( n = 20), minimal mitral regurgitation ( n = 13), and mild mitral regurgitation ( n = 5). Surface valve area has been calculated at 2.5 ± 0.4 cm 2 in partial homograft reconstruction and 2.7 ± 0.3 cm 2 in total homograft replacement, with a transvalvular gradient of 3 ± 4 mm Hg. Conclusion: In a selected group of patients, the use of mitral homografts significantly extended the present limitations of reparative surgery of the mitral valve. (J T HORAC C ARDIOVASC S URG 1996;111:367-80)