The Journal of Bone and Joint Surgery (American). 2008;90 Matthew B. Collier, MS1, C. Anderson Engh, Jr., MD1, Kyle M. Hatten, BA1, Stuart D. Ginn, BA1, Todd M. Sheils, MD1 and Gerard A. Engh, MD1 Background: Archived serial radiographs of knee replacements provide indirect evidence of the clinical performance of ultra-high molecular weight polyethylene tibial bearings. Our purpose was to determine the loss of thickness in polyethylene tibial inserts that were of the same design but had been sterilized differently. Methods: Four hundred and sixteen knees with an Anatomic Modular Knee primary total knee replacement had five to eighteen years of follow-up and a posterior cruciate ligament-retaining polyethylene tibial insert that had been sterilized with either gamma radiation in air (from 1987 to 1993) or gamma radiation in an inert gas (from 1993 to 1995), or by means of a nonirradiation gas-plasma method (from 1995 to 2001). Readings were taken from 1975 weight-bearing anteroposterior radiographs of the knee on which the top surface of the metal tibial tray had nearly projected as a line. In each tibiofemoral compartment, loss of polyethylene thickness was calculated as the standard thickness of the insert minus the measured magnification-corrected thickness. For each sterilization method, thickness loss was plotted versus the corresponding follow-up time, and a mixed linear regression model was used to quantify the change in thickness over time. Multiple linear regression analysis was used to determine whether loss of thickness was associated with implant factors, patient variables, and early postoperative limb alignment. Results: According to the mixed model regression coefficients, loss of medial compartment thickness was greatest for inserts that had been sterilized with gamma radiation in air (0.15 mm/yr), least for those sterilized with gamma radiation in an inert gas (0.02 mm/yr), and of an intermediate magnitude for those sterilized with a nonirradiation method (0.06 mm/year); loss of lateral compartment thickness ranged from 0.03 to 0.06 mm/yr. Notably, no insert irradiated in an inert gas showed a loss of thickness of >1 mm, and no nonirradiated insert showed a loss of >2 mm. Loss of medial compartment thickness in inserts sterilized with radiation in air increased significantly with a greater polyethylene shelf age, lesser patient age, and more varus alignment of the limb in the early postoperative period (p < 0.01). Conclusions: In this total knee design, tibial polyethylene performance improved markedly after discontinuation of the gamma radiation in air sterilization technique. Future concerns are that bearings sterilized with radiation in an inert gas may oxidize in vivo and develop fatigue wear because of free radicals generated during sterilization with radiation and that nonirradiated bearings may undergo greater losses in thickness from routine burnishing since they lack the cross-linking that accompanies sterilization with radiation. Level of Evidence: Therapeutic Level III. See Instructions to Authors for a complete description of levels of evidence.