From December 1969 to May 1975, 124 patients underwent aortic valve replacement with an allograft aortic valve sterilized by incubation in a low dose antibiotic solution and stored by refrigeration at 4 degrees C (4 degrees C stored valve group). From June 1975 to December 1987, 231 patients received an allograft aortic valve, sterilized by the same low dose antibiotic solution, but stored by cryopreservation in liquid nitrogen at -196 degrees C (cryopreserved valve group). The 4 degrees C stored valves were essentially nonviable, whereas the cryopreserved valves were viable at implantation. Of the 355 aortic valve replacements, associated procedures were performed in 127 patients. The 30-day mortality was 8.9% (confidence limits [C.L.] 6.2% ... 12.3%) (4 degrees C stored) and 4.8% (C.L. 3.3% ... 6.7%) (cryopreserved). Actuarial survival was similar in both groups, being 71% and 67% at 10 years in the 4 degrees C stored and cryopreserved valve groups, respectively (P = .18). The probability of a thromboembolic event was low, but appeared higher in the 4 degrees C stored valve group (actuarial freedom at 10 years, 90%) than the cryopreserved valve group (actuarial freedom at 10 years, 98%) (P = .01) probably related to associated mitral valve surgery. The actuarial freedom from allograft valve endocarditis at 10 years was 94% and 95% for the 4 degrees C stored and cryopreserved valve groups, respectively (P = .23). Reoperation was undertaken in 34 patients in the 4 degrees C stored group and 12 patients in the cryopreserved valve group for leaflet degeneration, endocarditis, or technical reasons.(ABSTRACT TRUNCATED AT 250 WORDS)