Purpose: The purpose of this study was to use a clinical simulation to determine whether wear of the internal surface of a titanium implant was greater following connection and loading of a one-piece zirconia implant abutment or a titanium implant abutment. Materials and Methods: Two implants received zirconia abutments and two received titanium abutments. The implants were secured into four fiber-reinforced epoxy resin disks that had been prepared to receive the internal-connection implants. The assemblies were cyclically loaded off-axis for a total of 1,000,000 cycles. At various intervals, the abutments were removed, photographed, examined using scanning electron microscopy (SEM), and returned to the implants for further testing. The area of titanium transfer from the implants to the abutments observed in the SEM images was quantified using image analysis software. Results: The method was able to quantify the area of material transferred to the abutments. There was considerably more wear associated with the zirconia abutments, but the rate of wear slowed after about 250,000 cycles. Parabolic curves were fit to the data. The projected mean ± standard deviation maximum area (wear) values associated with the titanium and zirconia abutments were 15.8 ± 3.3 × 103 µm2 and 131.8 ± 14.5 × 103 µm2, respectively, and this difference was statistically significant (P = .0081). Conclusions: The implants with the zirconia abutments showed a greater initial rate of wear and more total wear than the implants with the titanium abutments following cyclic loading. The amount of titanium transfer seen on the zirconia abutment increased with the number of loading cycles but appeared to be self-limiting. The clinical ramifications of this finding are unknown at this time; however, the potential for component loosening and subsequent fracture and/or the release of particulate titanium debris may be of concern. Keywords: cyclic loading, fatigue testing, fretting wear, implant abutment, implant-abutment interface, titanium, zirconia