Retrieval studies have shown that the interface between the ultra-high molecular weight polyethylene insert and metal tibial tray of fixed-bearing total knee replacement components can be a source of substantial amounts of wear debris due to fretting micromotion. We assessed fretting wear of polyethylene against metal as a function of metal surface finish, alloy, and micromotion amplitude, using a three-station pin-on-disc fretting wear simulator. Overall, the greatest reduction in polyethylene wear was achieved by highly polishing the metal surface. For example, highly polished titanium alloy surfaces produced nearly 20 times less polyethylene wear compared with blasted titanium alloy, whereas, decreasing the micromotion amplitude from 200 to 50microm produced approximately four times less polyethylene wear for the same blasted titanium alloy surface. Although the effect of the metal alloy was much smaller than the effect of metal surface roughness or the micromotion amplitude, CoCr discs produced slightly greater polyethylene fretting wear than titanium alloy discs under each condition. The results are essential in design and manufacturing decisions related to fixed-bearing total knee replacements.