The xenograft stentless valve was designed to emulate the haemodynamic performance of the allograft. Early outcomes using either surgical option (stentless xenograft valve or allograft) have been similar. However, follow-up outcomes remain to be compared. Between 1st January 1991 and 1st January 2001, 415 patients underwent aortic valve replacement. Two hundred and seventeen patients received an allograft and in 198 patients a Toronto stentless porcine valve (TSPV) was implanted. Mean time to follow-up was 6.3+/-4.4 years. Ten years freedom from structural valve deterioration (SVD) (TSPV 86+/-5%, allograft 82+/-5%, P=0.49) and freedom from reoperation (RE) (TSPV 80+/-4% vs. allograft 85+/-4%, P=0.61) were not significantly different. The TSPV was associated with significantly worse actuarial survival than the homograft (TSPV 40+/-4% vs. homograft 55+/-4%, P=0.02). However, after adjustment for other variables using a multivariate model, TSPV use was not an independent predictor of impaired late survival (LS) (P=0.44). Kaplan-Meier analysis in a subgroup of patients aged 45-65 years comparing LS, RE and SVD between xenografts and allografts identified similar results. The porcine stentless valve may be associated with similar clinical performance to the allograft over the medium to long-term.