Historically, hip joint simulators most often have been used to model wear of a bearing surface against a bearing surface. These simulators have provided highly accurate predictions of the in vivo wear of a broad spectrum of bearing materials, including cross-linked polyethylenes, metal-on-metal, ceramic-on-ceramic, and others in development. In recent years, more severe conditions have been successfully modeled, including jogging, stair climbing, ball-cup micro separation, third-body abrasion, and neck-socket impingement. These tests have served to identify improved materials and to eliminate some with inadequate wear resistance prior to their clinical use. Simulation of the knee joint is inherently more complex than it is for the hip. It is more difficult to compare the results of laboratory tests with actual clinical performance, due to the lack of accurate in vivo measures of wear. Nevertheless, knee simulators, based on force control or motion control, have successfully reproduced the type of surface damage that occurs in vivo (eg, burnishing, scratching, pitting) as well as the size and shapes of the resultant wear particles. Knee simulators have been used to compare molded versus machined polyethylene components, highly cross-linked polyethylenes, fixed versus mobile bearings, and oxidized zirconia and other materials, under optimal conditions as well as more severe wear modes, such as malalignment, higher loading and activity levels, and third-body roughening.