Objective To compare the biomechanical performance of Kangli hollow screws with sliding compression locking plate system (KHS) and conventional cannulated lag screws for fixation of type Pauwels III femoral neck fracture. Methods 7 cadaveric femurs were selected, vertical fractures (Pauwels III fracture, at 70° to the horizontal) were artificially conducted in these cadaveric proximal femurs by an orthopaedic surgeon and fixed by KHS screws with plate system or conventional cannulated lag screws. Samples were positioned at 75° of the femoral shaft to the horizontal, embedded in the mould and fixed in the experimental console. Optical sensors were set at the femoral neck around the osteotomy line. Then the loading were input in the vertical, horizontal lateral direction and rotating direction around the femoral neck axis, the maximal and minimal values between the fractured fragments and the corresponding values of the loading were recorded. The values of stiffness in three directions were calculated and compared. The CT data of the left femur of a 25 year-old healthy male volunteer was input into the corresponding software and vertical femoral neck fracture model was generated. Two finite element analysis models were obtained after the fracture being fixed using these two different implants, and the Von Mises stress distribution on the femur, implants and the interface between the fractured fragments and the relative motion between the fractured fragments were compared. Results In the vertical, horizontal lateral direction and rotating direction around the femoral neck axis, the stiffness of the KHS were 3 904±1 148 N/mm, 4 324±1 234 N/mm and 11.45±4.95 N·m/° respectively, higher than those of the CSs method with the values of 3 020±1 150 N/mm, 3 020±854 N/mm and 6.53±4.83 N·m/° respectively. The differences between the two groups were statistically significant (t=2.7194, 4.7694 and 2.9424; P=0.0347, 0.0050 and 0.0423). In the finite element analysis test, the maximal Von Mises stress values distributing on the femur and the screws in the KHS group were 40.1 MPa and 126.4 MPa, and those in the CSs group were 98.1 MPa and 145.5 MPa respectively, and both values of the former were lower than the latter. But the Von Mises stress value on the interface between the fractured fragments in the KHS group was 14.37 MPa, which was much higher than that in the CSs groupwhich was 9.39 MPa. The gap at the fracture site of the CSs fixation model was dramatically larger than that of KHS fixation model. Conclusion The KHS screws and plate system could provide better immobilization effect for vertical femoral neck fracture compared to the cannulated lag screws. The risk of the screws failure was lower and the fracture union would be easier to obtained by the fixation of KHS screws with plate system. Key words: Femoral neck fractures; Fracture fixation, internal; Biomechanics; Finite element analysis