Supracondylar fracture of the femur after total knee arthroplasty has an estimated frequency of 0.6%-2.5% among total knee recipients and presents an extremely difficult problem when encountered. The goal of this study is to determine the most stable method of fixation of these supracondylar fractures among currently available devices. Synthetic composite femurs with properties similar to human bone were used, and identical, unstable supracondylar fractures were created in each. Osteotomized specimens were placed into four groups of five. Each group was then tested with one of four devices: the Green-Seligson-Henry (GSH) intramedullary nail, AO 95 degrees blade plate, dynamic condylar screw and sideplate, and condylar buttress plate. After stabilization with the different types of fixation, the constructs were tested individually for bending stiffness in four modes: flexion, extension, varus, and valgus bending. The stiffest fixation was determined in each of the four bending planes. Resistance to all tested directions was greatest for the condylar screw and sideplate construct. Resistance to flexion (stiffness = 30.96 N/mm), extension (stiffness = 36.36 N/mm), varus (stiffness = 35.46 N/mm), and valgus forces (stiffness = 32.26 N/mm) was highest in the group fixed with the dynamic condylar screw. This may be due to the purchase gained by the large lag screw into the distal femur, or it may be the result of the total rigidity of the implant. Although the femoral samples used in this study do not duplicate the typical osteopenic bone encountered at the site of a total knee arthroplasty, they do allow direct comparison of the fixation devices by removing the variability associated with cadaveric bone samples.