The objective of this study was to determine the effect of screws and keel size on the fixation of an all-metal glenoid component. A prototype stainless-steel glenoid component was designed and implanted in 10 cadaveric scapulae. A testing apparatus capable of producing a loading vector at various angles, magnitudes, and directions was used. The independent variables included six directions and three angles of joint load, and five fixation modalities-three different-sized cross-keels (small, medium, and large), screws, and bone cement. Implant micromotion relative to bone was measured by four displacement transducers at the superior, inferior, anterior, and posterior sites. The components displayed a consistent response to loading of ipsilateral compression and contralateral distraction. Use of progressively larger keels did not significantly improve implant stability. Stability decreased as the angle of load application increased (P <.05). Screw and cement fixation resulted in the most stable fixation (P <.05).