Simulator wear tests were conducted for 1.5 million cycles on second generation metal-on-metal hip implants fabricated from cobalt-based alloy (ASTM F 1537-94) to compare their performance with each other and with both custom-made implants and retrieval studies published in the literature. In addition, factors controlling wear were explored. Six commercially available implants were tested (2 low carbon and 4 high carbon) for 1.5 million cycles in a MATCO™ hip simulator following previously established protocols. Volumetric wear was in the same range for all of the implants tested and much less than would be expected in comparable tests with conventional polyethylene cups. The wear of the commercial implants used in the present study was very similar to custom-made hip implants of similar diameter, clearance and surface roughness. Three of the high carbon implants of the present study had somewhat lower wear than expected when their surface roughness was taken into account but, in general, clearance, roughness and lambda (a ratio of lubricant film thickness to combined surface roughness) were factors controlling wear. The simulator wear of the present study showed some agreement with the in vivo wear estimated in retrieval studies in the literature.