Forty-four asymptomatic patients were catheterised at Boucicaut Hospital 9.5 months after aortic valve replacement to assess the haemodynamic performances of 21 Hancock pericardial (HP) and 23 porcine Carpentier-Edwards (CE) (standard model) bioprostheses, implanted in the aortic position. Left heart catheterisation was performed from a femoral approach; the simultaneous gradient was measured by planimetry and the functional valve surface area calculated at rest and after exercise. The resting calculated surface area of the HP was greater than that of the CE bioprostheses (equation: see text). The transvalvular pressure gradient was lower in the HP than in the CE group (7.8 +/- 4 vs 15.3 cf243 6 mmHg; p less than 0.005). After exercise (15 patients) the calculated surface area increased with the increased transvalvular blood flow in both groups but at each flow rate the calculated valve surface area was greater in the HP group. The haemodynamic performance of the CE bioprosthesis is inferior to that of the HP bioprosthesis, especially in the smaller models. However, the CE bioprosthesis would seem to be mechanically more reliable in the long term than the HP bioprosthesis has since been withdrawn from the market.