Objective:To compare biomechanical stabilities between screw-plate fixation and non-cannulated screw fixation for Lisfranc ligament injury by a 3-D finite element analysis.Methods:A 3-D model of a healthy foot was developed from computed tomography images. The 1st and the 2nd dorsal tarsometatarsal ligaments and Lisfranc ligament were cut in the 3-D model of a healthy foot to establish a Lisfranc ligament injury model, in which screw-plate fixation (with 2 locking plates and 8 standard screws and one non-cannulated screw) and non-cannulated screw fixation (with 3 non-cannulated screws) for Lisfranc ligament injury were simulated respectively. Finite element analyses were carried out by Abaqus 6.14 software after loads were added in the 3-D models of screw-plate fixation and non-cannulated screw fixation for Lisfranc ligament injury. The overall stress-strain nephogram, the stress distribution and displacement of the foot bone, and the stress distribution on the internal fixation system were compared between the 2 kinds of models.Results:Under the same load, the stress of the whole screw-plate fixation was concentrated on the fixators, and the stress of the non-cannulated screws was also greater than that of the bones. In both models, the strain of the whole foot led to arch collapse, especially in the medial column. The maximum stress on the screw holes in the medial and middle columns in the screw-plate fixation model was 39.91 MPa, smaller than that in the non-cannulated screw fixation model (53.13 MPa). The relative displacement of the first metatarsal joint in the screw-plate fixation model was 8.515 × 10 -1 mm, much greater than that in the non-cannulated screw fixation model (3.893 × 10 -1 mm). Stress concentration was observed in both models. The stress of the screw-plate system was concentrated on the fibular side of the middle section of the plate used to fix the first tarsometatarsal joint, decreasing towards both ends. The maximum stress of the non-cannulated screws was located in the middle of the medial column screw for fixation of the first tarsometatarsal joint, significantly greater than those of the both ends. The maximum stress of the screw-plate system was 239.5 MPa, smaller than that of the non-cannulated screws (256.8 MPa). Conclusions:Non-cannulated screw fixation demonstrates a greater biomechanical stability for Lisfranc ligament injury than screw-plate fixation. However, the former may have a higher risk of screw breakage because it bears a greater stress.