The effect of manufacturing process on the wear and mechanical performance of a total knee replacement (TKR) design was investigated with the use of a force-controlled knee joint simulator. Ultra-high molecular weight polyethylene (UHMWPE) tibial inserts processed by direct compression molding from 1900H resin were compared to UHMWPE tibial inserts machined from a compression-molded sheet of GUR 1050. Both sets of components had the same posterior-cruciate-retaining geometry, and were identically aligned with cobalt-chromium-molybdenum alloy femoral components. Wear tests were conducted at a frequency of 1 Hz for 4 million cycles with the use of a standard walking cycle pattern. Implant kinematics, including anterior-posterior (AP) displacement and internal-external (IE) rotation in response to applied loads were monitored. Gravimetric wear, surface roughness, and surface morphology were used to characterize the wear process of the UHMWPE inserts. Results showed that the molded UHMWPE inserts exhibited less gravimetric wear over time than the machined inserts of the same design. Both the machined and molded components exhibited scratching, pitting, and burnishing over their wear areas. The AP displacement distance per cycle of the molded tibial inserts decreased over the course of testing, resulting in a shorter total testing displacement for this group compared to machined tibial inserts. Although AP displacement distance per cycle for machined tibial inserts did not change significantly over the course of testing, their position relative to the femoral components shifted posteriorly over time, resulting in an elongated wear track. (C) 2001 John Wiley & Sons, Inc.