Background Surgeons can control not only the angle but also the depth of suture anchor placement during arthroscopic rotator cuff repair, although the tendency may be to place suture anchors on the deep side to avoid damage from prominent anchor eyelets. However, little information is available regarding possible effects of suture anchor depth on construct failure mechanisms. Hypothesis Anchor depth affects the mode of suture failure with physiologically relevant cyclic loads. Study Design Controlled laboratory study. Methods Metallic screw-in suture anchors loaded with No. 2 braided polyester sutures were inserted into the bovine infra-spinatus footprint with the eyelet proud, standard, or deep. Sutures were hand tied to create a closed loop. Constructs were cyclically loaded from 10 to 90 N and, if still intact at 500 cycles, taken to ultimate failure (maximum load). Results When clinical failure was defined as greater than 3-mm construct elongation, anchors placed with the eyelet deep experienced statistically earlier clinical failure via cutting of the suture through the bone (P <. 02). However, anchors placed at this level did not experience catastrophic failure during cyclic loading. The standard and proud anchors experienced 3 mm of elongation at a greater number of cycles, but the suture material degraded at the anchor eyelet, and a majority of these constructs broke during cyclic physiologic loading. At failure testing, the deep anchors had a significantly increased failure load (164 N) compared to standard (133 N) (P <. 04) and proud (113 N) anchors (P <. 005). Conclusion Varying the depth of suture anchor insertion changes the mechanical properties and mode of failure of suture anchor constructs. Clinical Relevance Surgeons should be aware of the effects of suture anchor depth and abrasive eyelet wear on construct failure during arthroscopic rotator cuff repair.