To compare prosthetic valve behaviour with that of the natural aortic valve, experiments were performed in vitro as well as in vivo. In a mock circulation system, cinematographic high-speed recordings of the valvular behaviour were made for Björk-Shiley, St Jude and Hancock heart valve prostheses. Simultaneously with the film recording, the aortic flow and the left ventricular and aortic pressures were measured. The closing behaviour of the natural aortic valve was recorded in in vivo experiments following the same measuring technique. Comparison of the film frames with the aortic flow signal revealed that the mechanical prostheses mainly close due to the back flow in the early phase of diastole; they close only for 5% of their cross-sectional area during systolic ejection. The Hancock bioprosthesis closes already for 45% during the flow deceleration phase of systole, which is however significantly less than the corresponding closure of the aortic valve as recorded in vivo (74%). The change in fluid viscosity from 3.10-3 Ns/m2 to 1.10-3 Ns/m2 does not affect the closing behaviour of the mechnaical prostheses. Reducing the peak value of systolic aortic flow by a factor two decreases the cross-sectional area of the prosthetic valves at peak systole. When the natural aortic valve closing mechanism is applied to a disc-valve prosthesis, a partial (30%) systolic valve closure is found.