The ultimate pullout strength and fatigue properties of a screw-design suture anchor implanted in the anterior glenoid rim were investigated and compared with results from a nonscrew-design suture anchor. Twenty-two cadaveric glenoids were harvested and one to two anchors were implanted in the superior and inferior quadrants. Fifty-seven Statak 3.5 anchors (Zimmer, Warsaw, IN) were tested and compared with results obtained in a previous study on 50 Mitek GII anchors (Mitek Products, Inc, Westwood, MA). The specimens were mounted on an Instron fatigue testing machine (Instron Corp, Canton, MA) and cycled between preselected minimum and maximum loads until pullout. The Mitek GII maintained a higher pullout strength than the Statak 3.5 after cyclic loading. Cortical thickness at the implantation sites was measured, and found to decrease monotonically from superior to inferior positions. The ultimate pullout strength, and subsequently the fatigue life, of both types of suture anchors depended directly on cortical thickness. The significantly lower performance of both anchors when placed inferiorly emphasizes the importance of correct anchor selection, number, and placement in this region. All anchors settled during the first 10 to 100 cycles, resulting in partial exposure of the implant. Intraoperative cycling of the anchors before suture tying may be necessary to achieve complete settling and prevent subsequent loss of coaptation between capsule and glenoid. The study shows that for the anchors to last 1,000 cycles or more, less than 50% of the theoretical ultimate pullout strength should be applied cyclically. With aggressive early rehabilitation exercises, this significant decrease in fixation strength could shift reconstruction failure from suture breakage or soft tissue tearing to anchor pullout.