Studies have shown that titanium implants can be challenging to explant due to the material's excellent biocompatibility and resulting osseointegration. Clinically, titanium alloy nail interlocking screws may require removal to dynamize a construct or revise the nail due to nonunion, infection, pain, or periprosthetic fracture. This study was designed to determine what variables influence the removal torque for titanium alloy interlocking screws. An intramedullary nail with four interlocking screws was used to stabilize a 1-cm segmental femoral defect in a canine model for 16 weeks. The animals were observed to be active following a several-day recovery after surgery. In six animals, the femora and implanted nail/screws were first tested to failure in torsion to simulate periprosthetic fracture of an implant after which the screws were then removed. In four additional animals, the screws were removed without mechanical testing. Both intraoperative insertional and extraction torques were recorded for all screws. Mechanical testing to failure broke 10/24 screws. On average, the intact screws required 70% of the insertional torque during removal while broken screws only required 16% of the insertional torque (p < 0.001). In addition, intact screws closer to the fracture required 2.8 times more removal torque than the outboard distal screw (p < 0.005). On average, the angle of rotation to peak torque was ∼80°. The peak axial load did not significantly correlate with the torque required to remove the screws. On average, the removal torque was lower than at the time of insertion, and less torque was required to remove broken screws and screws remote to the fracture. However, broken screws will require additional time to retrieve the remaining screw fragment. This study suggests that broken screws and screws in prematurely active patients will require less torque to remove.