Alumina-on-alumina hip implants with microseparation were run in a hip simulator for comparison of the nonseparation simulator mode and retrievals. The 28-, 32-, and 36-mm Biolox-forte implants were run to 5 million cycles with the use of 50% newborn calf serum. Howmedica Osteonics Trident cups with titanium backing were used in all sets. In standard (STD) and microseparation (MSX) mode, the typical biphasic wear trend was evident, but the MSX test mode had much higher magnitudes. There was a 5-fold increase for run-in wear and up to a 35-fold increase in steady-state wear. The stripe wear on the ball formed early, but did not progress in grade beyond 0.6 Mc. The locations of the stripes were similar in retrieved and simulator balls. However, the stripes from the simulator were narrower than short-term retrievals and much narrower than some long-term retrievals. The long-term retrieved balls had a grade of wear greater than the simulators. In vivo a broader range of motion occurs and this may lead to the wider stripe observed on the retrievals. These observations suggested that simulators could produce the loading and kinematics similar to a patient walking but not necessarily the variety of motions possible in the in vivo situation.