The initial stability of femoral neck osteosynthesis provided by three self-reinforced poly-L-lactide (SR-PLLA) or three metallic cannulated screws was compared. A standard transverse subcapital osteotomy was created with a hand saw in 19 pairs of human cadaver femora. Two fixation methods were randomly used in each pair. Fixations were exposed to a progressive cyclic loading test to determine the deflection curves, the ultimate load-carrying capacity, and the total amount of load absorbed. One pair served as a pilot and two nonoperated pairs served as controls. The positioning of the three fixation screws, whether one screw proximally and two screws distally or vice versa, provided equal stability. Metallic fixation proved to be more stable than SR-PLLA fixation as the average maximum load-carrying capacity was 3400N for the metallic and 2600N for the SR-PLLA fixation. However, SR-PLLA screws showed sufficient fixation properties for considering clinical trials in Garden I-II femoral neck fractures in well cooperative patients.