A number of recent reports have suggested that valve design has an influence on long-term survival after aortic valve replacement (AVR). The present study examines the clinical results of two large multicenter trials of the Hancock II stented and Medtronic Freestyle stentless bioprosthetic valves. In all cases a subcoronary implant technique was used. All patients were operated on between 1991 and 1994, and all patients were monitored for 5 years. The 5-year actuarial survival rate was 86% for Freestyle and 77% for Hancock patients (P<.05). Late mortality occurred in 40 Freestyle patients (8.8%) compared with 36 Hancock II patients (16.1%, P = .0074). Cox's proportional hazard models were used to identify independent predictors of 5-year survival for the two groups. Hazard ratio (HR) and P value are presented in parentheses. The final analysis included the interactive variable valve*age. The resultant model indicates that the interaction of age and valve type has a significant effect on outcome. In patients < or =60 years of age at the time of operation, AVR with a Hancock II valves was associated with nearly a fivefold risk of death compared with patients of the same age who received a Freestyle valve (HR = 4.97, P = .0004). However, this survival advantage decreased by 50% with each decade thereafter (valve*age interaction HR = 0.50, P = .0027). Our data indicate that AVR with a stentless valve confers a survival advantage to the patient. The probability of dying within 5 years of surgery is 50% greater in patients who received a Hancock II versus a Freestyle valve (HR = 1.50, P = .0442). However, this survival advantage is most prominent in younger patients (<60 years of age) where the probability of death is fivefold greater with Hancock than Freestyle valves. With advancing age the benefits of stentless valves are diminished.