OBJECTIVES: The number of locking screws required per fragment during bridging osteosynthesis in the dog has not been determined. The purpose of this study was to assess the survival of two constructs, with either two or three screws per fragment, under cyclic torsion.METHODS: Ten-hole 3.5 mm stainless steel locking compression plates (LCP) were fixed 1 mm away from bone surrogates with a fracture gap of 47 mm using two bicortical locking screws (10 constructs) or three bicortical locking screws (10 constructs) per fragment, placed at the extremities of each LCP. Constructs were tested in cyclic torsion (range: 0 to +0.218 rad) until failure.RESULTS: The 3-screws constructs (29.65 ± 1.89 N.m/rad) were stiffer than the 2-screws constructs (23.73 ± 0.87 N.m/rad), and therefore, were subjected to a greater torque during cycling (6.05 ± 1.33 N.m and 4.88 ± 1.14 N.m respectively). The 3-screws constructs sustained a significantly greater number of cycles (20,700 ± 5,735 cycles) than the 2-screws constructs (15,600 ± 5,272 cycles). In most constructs, failure was due to screw damage at the junction of the shaft and head. The remaining constructs failed because of screw head unlocking, sometimes due to incomplete seating of the screw head prior to testing.CLINICAL SIGNIFICANCE: Omitting the third innermost locking screw during bridging osteosynthesis led to a reduction in fatigue life of 25% and construct stiffness by 20%. Fracture of the screws is believed to occur sequentially, starting with the innermost screw that initially shields the other screws.