For patients suffering from arthritic conditions, the dominant type of bearing replacement used in artificial hip joints has been a metal ball running in a hemispherical plastic cup. It has been hypothesized that a major cause of accelerated wear with this combination is a 3rd-body wear mechanism, interacting between the CoCr counterfaces and conventionally crosslinked polyethylene (CXPE) bearings. In addition, over the past 30 years the dominant form of implant fixation to bone has been a polymethyl-methacrylate (PMMA) bone cement. As such, the potentially abrasive wear mechanism may be related to the presence of circulating, micron-sized, particulates of PMMA. One laboratory wear study indicated an unexpected type of roughening mechanism for such bearings in that, compared to ‘clean’ lubricants, wear of CXPE and HXPE cups in abrasive slurry increased 6- and 20-fold, respectively. However the surprising finding was that there was no corresponding change in CoCr roughness. The abrasive mechanism appeared to be a transient formation of PMMA layers on the CoCr counterfaces. Modern cup designs predominantly use highly crosslinked polyethylenes (HXPE), which have a much short clinical history. Thus it is conceivable that wear of the HXPE bearings will also be influenced both by circulating PMMA debris (3rd-body wear) and damaged CoCr counterfaces (2-body wear). The present study re-examined the effects of 3rd-body PMMA particulates on the wear of CXPE and HXPE cups using both 32 and 44 mm diameter CoCr and ceramic balls. The 32 mm CXPE represented the control bearing. The hip simulator wear study was run in two phases to compare the effects of abrasive wear. Conventional polyethylene cups (CXPE: 35 kGy irradiated) and cups highly crosslinked by 75 kGy irradiation (HXPE) typical of modern bearings were run with both CoCr and Al 2O 3 balls as the mating counterfaces. The polyethylene cups were mounted in “inverted” position for the debris studies in a multi-station hip simulator. In phase I, all cups were run in ‘clean’ serum lubricant for 1.5 million load cycles (1.5 Mc). In phase II, the cups were run in an abrasive slurry for 2 Mc duration. A commercial PMMA bone cement powder was used as the “abrasive” medium and the PMMA slurry (5 mg/ml concentration) was added to all lubricant chambers at multiple test durations. Wears assessed by the average bearing weight loss and characterized by linear regression in each test phase. Wear and roughness effects on polyethylene, CoCr and Al 2O 3 bearings was analyzed by reflected light microscopy, white light interferometry and scanning electron microscopy. The total wear study was run to 3.5 Mc duration. Compared to phase-1 wear of control CXPE [32 mm] combinations, the wear of ‘new’ HXPE [44 mm] cups was a 6-fold lower. In the phase-2 study, HXPE cups showed different trends depending on whether CoCr or Al 2O 3 balls were used. Thus the wear of HXPE [44C] decreased by 43%, whereas the wear of HXPE [44M] increased 500%. The wear of the “pre-worn (cup after 10 million cycles of simulator test)” HXPE [44M] was 4-folds higher than the ‘new’ HXPE [44M] cups. In phase-2, the controls averaged 320% higher than the new HXPE [44 mm]. The wear of control CXPE was directly proportional to debris concentration, i.e. with a 50% reduction in slurry concentration, there was a corresponding reduction wear. It was noted that HXPE wear increased only 5-fold compared to the 80-fold previously measured when using double the slurry concentration. This may represent the sensitivity of HXPE to such abrasive wear conditions under the same experimental conditions. The HXPE bearings maintained a superior finish to CXPE (controls). In phase I, the latter's average roughness decreased from 280 to 150 nm whereas ‘pre-worn’ HXPE cup decreased from 180 to 40 nm. During phase-2, the CXPE and HXPE with CoCr bearings averaged 7 and 5 times rougher than with Al 2O 3 combination. A significant correlation was found between phase-2 wear magnitude and resulting cup roughness. Stepwise multiple regression analysis revealed that cup roughness showed the strongest relationship to wear magnitude. Ball roughness was not a factor since CoCr and Al 2O 3 balls maintained a high quality finish (Sa < 12 nm). However, during Phase-2, SEM and interferometry revealed numerous PMMA particles adhering to uncleaned CoCr counterfaces. In contrast the Al 2O 3 surfaces were relatively free of PMMA contamination. Thus use of Al 2O 3 balls minimized the risk of both 2-body and 3rd-body PMMA abrasive wear. We described for the first time that the average cup roughness (Sa) were directly proportional to the wear rate. Therefore the surface finish of the bearings (CoCr, CXPE and HXPE) appears to be an important aspect of the hip wear mechanism.