The design of artificial heart valves has traditionally been based on the development of a prototype device which was then subjected to extensive laboratory testing in order to confirm its suitability for clinical use. In the past the in vitro assessment of a valve's performance was based principally on the measurement of parameters such as pressure difference, regurgitation and, more recently, energy losses. Such measurements can be defined as being at the 'macro' level and rarely show any clinically significant differences amongst currently available prostheses. The analytical approach to flow through heart valves has previously been hampered by difficulties experienced in solving the relevant equations of flow particularly in the case of pulsatile conditions. Computational techniques are now available which enable appropriate solutions to be obtained for these problems and consequently provide an opportunity for detailed examination of the 'micro' level of flow disturbances exhibited by the different valves. This present preliminary study is designed to illustrate the use of such an analytical approach to the flow through prosthetic valves. A single topic has been selected for this purpose which is the comparative value of steady versus pulsatile flow testing. A bileaflet valve was chosen for the analysis and a mathematical model of this valve in the aortic position of the Sheffield Pulse Duplicator was created. The theoretical analysis was carried out using a commercially available Computational Fluid Dynamics package, namely, FIDAP, on a SUN MICROSYSTEMS 10-30 workstation.(ABSTRACT TRUNCATED AT 250 WORDS)