OBJECTIVE The study aim was to define the long-term results after aortic valve replacement (AVR) with freehand stentless xenografts. METHODS A retrospective analysis was performed of 376 consecutive patients (195 males, 181 females; mean age 70 +/- 7 years; range: 26-87 years) who underwent stentless AVR between October 1992 and April 2000. In total, 164 patients received Toronto SPV, 106 Biocor PSB, 101 Cryolife-O'Brien, and five other valves. Indication for surgery was valve stenosis (n = 246), regurgitation (n = 50), mixed lesion (n = 75) and prosthetic valve failure (n = 5). Mean preoperative NYHA functional class was 2.9 +/- 0.6 (range: 2-4). Associated procedures were required in 153 patients (41%), including coronary artery bypass grafting (n = 97), mitral operation (n = 26), ascending aortic replacement (n = 17) and 'other' (n = 13). Mean valve size was 25 +/- 3 mm (range: 21-29 mm), mean aortic cross-clamp time was 96 +/- 23 min (range: 42-186 min), and mean cardiopulmonary bypass time 128 +/- 34 min (range: 65-363 min). RESULTS Ten patients (2.7%) died in hospital, due to cardiac causes (n = 6), cerebrovascular accident (n = 3) and multiorgan failure (n = 1). During a mean follow up of 40 +/- 20 months (range: 1-90 months) there were 33 late deaths. Survival was 96 +/- 1%, 83 +/- 3% and 80 +/- 4% and valve-related event-free survival 97 +/- 1%, 87 +/- 2% and 82 +/- 6% at one, five and seven years, respectively. Valve failure occurred in 21 (6%) patients (10 structural, nine non-structural, two endocarditis): freedom from structural valve deterioration was 99 +/- 1%, 95 +/- 2% and 92 +/- 4% at one, five and seven years. Reoperation on the xenograft was required in 17 (5%) patients, giving a freedom from reoperation of 99 +/- 1%, 94 +/- 2% and 90 +/- 4% at one, five and seven years. Freedom from valve failure and reoperation were significantly lower with Cryolife-O'Brien valves, but freedom from structural valve failure was similar. The average NYHA class was improved at follow up (1.4 +/- 0.7 versus 2.9 +/- 0.6 preoperatively, p = 0.01). CONCLUSIONS Long-term survival and freedom from valve-related adverse events after stentless AVR in an elderly population were satisfactory. Freedom from structural deterioration was excellent and comparable for all xenograft models, while non-structural deterioration may have been more common with one valve model. Despite the advanced patient age, functional improvement late after operation may be expected.