Objective To investigate the influence of displacement of femoral and tibial components on the biomechanics of femoral or tibial bone in coronal view. Methods A series of CT and MRI of the left knee joint of a Han male volunteer was taken and a three-dimensional finite element model of the healthy knee joint was established. The femoral component and the tibial component were designed with varus 6°, varus 3°, 0°, valgus 3°, and valgus 6°, and were combined into 25 three-dimensional finite element model (FEM) of medial unicompartmental knee arthroplasty. A 1 000 N load was applied along the femoral mechanical axis. The von Mises cloud stress distribution was observed. Moreover, the lateral compartment load ratio, the high contact stress of cancellous bone and medial cortical bone below the tibial component, the upper surface of the polyethylene liner, and the femoral cartilage in the lateral compartment was measured. The statistically significant indicators compared with the neutral position (0° varus or valgus of the tibia and the femoral prosthesis, and 5° posterior slope of tibia prosthesis) were identified by scatter plots to find the dense and sparse areas of point items. The optimal position of the femoral component and the tibial component was determined by the number of items with statistical significance in the sparse area. Results When the femoral component was placed at 0° position, there was no significant difference in the high contact stress of cancellous bone below the tibial component in the five groups. When the femoral component was placed at 0° position, the tibial component was 6° varus or 6° valgus and the stress was increased by 9.21±3.38 MPa and 9.08±4.13 MPa (P<0.05), respectively. With the changes of femoral and tibial components from 6° varus to 6° valgus, the high contact stress of the medial cortical bone below the tibia was gradually decreased (P<0.05). When the femoral component was placed at 0°, the tibial component changes from 6° varus to 6° valgus without significant difference in the high contact stress on the upper surface of each group of polyethylene gasket. Compared with the neutral position group, the high contact stress of the 6° varus or 6° valgus group were increased by 2.88±2.53 MPa and 3.47±2.86 MPa, respective ly (P<0.05). The lateral compartment load ratio and the high contact stress of lateral compartment femoral cartilage was gradually decreased (P<0.05), when the femoral and tibial components changed from 6° varus to 6° valgus. The number (2.8%, 1/36) of indicators in the sparse area (the combination of all combinations of femur or tibia from 3° varus to 3° valgus) was less than that (57.8%, 37/64) in the dense area (set of all combinations except sparse area), and the difference was significant (χ2=29.61, P<0.001). Conclusion It is suggested that the position of the femoral component and the tibial component in fixed medial unicompartmental arthroplasty should not exceed 3° varus or valgus in patients with standard lower limb alignment. Key words: Osteoarthritis, knee; Arthroplasty, replacement, knee; Prosthesis fitting; Finite element analysis