The mechanical complications of osteosyntheses after hip fractures are previously investigated by mostly static or dynamic uniaxial loading test systems. However, the physiologic loading of the hip joint during a normal gait is a multiplanar, dynamic movement. Therefore, we constructed a system to test osteosyntheses for hip fractures under physiologic multiplanar loading representative of normal gait. To evaluate the testing system, 12 femora pairs were tested under 25,000 cycles with two standard osteosyntheses (Proximal Femoral Nail Antirotation/Gamma3 Nail). For angular movement, the varus collapse to cut out (proportional to(CO)) (proportional to(CO) = 4.8 degrees +/- 2.1 degrees for blade and proportional to(CO) = 7.8 degrees +/- 3.8 degrees for screw) was the dominant failure mode, and only slight rotational angle shifts (proportional to(Rot)) (proportional to(Rot) = 1.7 degrees +/- 0.4 degrees for blade and proportional to(Rot) = 2.4 degrees +/- 0.3 degrees for screw) of the femoral head around the implant axis were observed. Angular displacements in varus direction and rotation were higher in specimens reinforced with screws. Hence, the cut out model and the migration directions showed a distinction between helical blade and hip screw. However, there were no significant differences between the different implants. The new setup is able to create clinical failures and allows to give evidence about the anchorage stability of different implant types under dynamic gait motion pattern.