Background and hypothesis Posterior glenoid bone loss is found in a majority of patients with advanced osteoarthritis of the shoulder. In total shoulder arthroplasty, several methods currently exist for management of this bone loss, including the use of an augmented glenoid component. Different augmented glenoid designs would be expected to vary in their resistance to loosening during mechanical bench-top testing. Our hypothesis is that a stepped augmented glenoid component will have less mechanical liftoff than augmented components of varying designs without a step. Materials and methods Four glenoid prototypes articulated with a humeral head were loaded with a 170-lb compressive load and with 4 mm of posterior-superior translation of the humeral head to 100,000 cycles. Anterior glenoid liftoff was measured. Results The stepped glenoid component had significantly lower liftoff values (P < .05) compared with several other designs at many of the test intervals. Discussion A stepped design for an augmented glenoid component has superior fixation and less anterior glenoid liftoff in the presence of eccentric loading and may have better long-term clinical results.