OBJECTIVE Mechanical wear is an important consideration for the Björk-Shiley Delrin (BSD) heart valve, the disc of which is periodically impacted against the inlet strut by the momentum of the blood flow during closure. Impact wear theory was used in designing experiments as well as establishing theoretical evaluation and projections of wear life. METHODS The experimental apparatus involved a pivotal hammer device where the striking face could be varied by the inclusion of distinct spherical shapes; the sharper the radius, the higher the contact stress induced. The striking speed could be varied between one and three m/s, and the repetition rate between 2-50 Hz. Blood was simulated with an aqueous 42% glycerol solution. The same wear mechanism was induced on Delrin occluder disc specimens as the one governing the wear observed in Delrin occluder discs of some of the explanted BSD heart valves; thus two wear parameters, c and g, were established for the prediction of wear in BSD implants. The experimental runs were repeated on as many as four simultaneous specimens, for up to 7.5 x 10(7) cycles. RESULTS The wear history was found to comprise three distinct regions, corresponding to (a) initial plastic deformations; (b) zero wear; and (c) measurable wear. The zero wear region produced very little change in the contact dimensions, and represented a fatigue threshold to the progress of a wear scar. Procedures for analytical establishment of the zero wear limit are outlined in the paper. Measurable wear was governed by the process of increasing conformance between the repetitively contacting surfaces. The differential equation of wear was then used with the physical parameters c and g previously determined experimentally. CONCLUSIONS The maximum depth of the wear scar after cycling the equivalent of 20 years was predicted to be 278 microns. This value appears to represent a safe dimension for BSD heart valves.