Unstented homograft valves offer several theoretical advantages when used for patients who have had previous operations on the aortic valve. Between January 1970 and February 1993, 177 patients received unstented homograft valves after previous aortic valve operations. One hundred thirty-four patients had previous aortic valve replacement in the form of homografts (101 patients), mechanical prostheses (24 patients), and bioprostheses (9 patients), and 43 had previous valve repair. The indication for reoperation was deterioration of a noninfected valve (124 patients), infective endocarditis (40 patients), and failure of a noninfected mechanical valve (12 patients). Fresh homograft valves were implanted in 60 patients, homografts preserved in antibiotics were used in 111 patients, and 6 patients received cryopreserved valves. Aortic valve and root replacement was performed in 60 patients, and in 117 the homograft was inserted freehand in the subcoronary position. The early mortality was 5.1%. The actuarial survival at 10 years was 71%. Multivariate analysis demonstrated that patients with previous homograft replacement have a better long-term survival than patients who had previous mechanical valves (p = 0.017). The freedom from valve-related death and reoperation was 70% at 10 years. Fresh homografts faired better than antibiotic-sterilized homografts (p = 0.007). None of the patients had recurrence of endocarditis at 6 months, although 1 patient died of uncontrolled infection despite valve replacement. The freedom from recurrent endocarditis was 88% at 10 years. We conclude that unstented aortic homografts provide good early and long-term results for aortic valve reoperations, particularly in patients with previous homograft replacement. Recurrent endocarditis is uncommon even in patients operated on for prosthetic valve infections.