From July 1992 to October 1994, we inserted new Toronto SPV stentless aortic heterografts (SJM Med. Inc., St Paul, Minneapolis, USA) in 40 of a series of 50 consecutive patients older than 70 years. The mean age was 75.7 years (range 70 - 86 years). All, except 4 patients, were pre-operatively in NYHA functional class III or higher. The aortic clamp time was significantly higher in the stentless groups (75 v. 53 minutes, P < 0.001). The average 25.5 mm size of the implanted valves stands in stark contract to the low body surface area (1.69 m2) of this patient groups. The surgeon's (in)experience was the major reason for the drawbacks (5/50) associated with a stentless procedures. The follow-up period ranged from 2 to 27 months and was complete in 100% of cases. We encountered 1 hospital death and no late deaths (97.5% actuarial survival). The mean NYHA class at follow-up was 1.5, and without exception patients were in class I or II. We noted one transient ischaemic attack immediately postoperatively and another later incident in a patient with a previous severe vascular history. With a low-intensity anticoagulation regiment for the first 3 months, there were two incidents of haemorrhaging necessitating premature anticoagulation withdrawal. Echocardiographic transthoracic valvular gradients compared favourably with the reported gradients of other biological valves, especially the smaller ones the significantly better haemodynamics were noted in most cases 6 months after implantation. Comparison of data with stented valves implanted during the same period indicates that the average size of the stentless valves was significantly higher (22.3 v. 25.5 mm, P < 0.001) in an equivalent population.