Objective: To determine the failure method of simulated equine medial femoral condyle (MFC) subchondral bone defects under compression and the influence of screw placement on failure resistance.Study Design: In vitro study. Sample Population: Composite disks (CD) simulating the moduli of yearling bone in the MFC. Methods: Four CD conditions were tested, all with a 12.7 mm void (n = 6 per condition): intact (no void), void only, void with a 4.5 mm screw placed in neutral fashion, and void with a 4.5 mm screw placed in lag fashion. Composite disks of each condition were tested under monotonic compression to 6000 N and cyclic compression to 10 000 cycles. Observable failure, load at first observable failure, and displacement at peak 2000 N load were compared among conditions.Results: Specimens failed by cracking at the superior aspect of the void or the screw exit hole. After monotonic loading, cracks were observed 6/6 CD with a void, 6/6 CD with a void/lag screw, and 5/6 CD with a void/neutral screw. After cyclical testing, cracks were noted only on the superior aspect of 6/6 CD with a void and 3/6 CD with a void/lag screw. Displacement at peak load was 0.06 mm (intact), 0.32 mm (void), 0.24 mm (void/lag screw), and 0.11 mm (void/neutral screw). Conclusion: Model MFC voids failed by superior cracking that was resisted by lag and neutral screw placement.Clinical Significance: Neutral screws may be an acceptable treatment for subchondral lucencies in the MFC.