Objective To investigate biomechanical differences of two posterior occipitocervical internal fixation techniques for treating basilar invagination with atlantoaxial dislocation (BI-AAD). Methods Intra-articular cage + posterior occipital plate+C2 pedicle screw (Cage+C2PS+OP), and intra-articular cage+C1 lateral mass screw+C2PS (Cage+C1LMS+C2PS) models were established based on occipitocervical CT data of the BI-AAD and clinical operation scheme, and the stability of atlantoaxial joint and stress distribution characteristics of C2 endplate and implanted instruments under different motion states were analyzed. Results Compared with the Cage+C1LMS+C2PS model, the atlantoaxial range of motion ( ROM) under flexion, extension, lateral bending and axial rotation in the Cage+C2PS+OP model were reduced by 5. 26% , 33. 33% , 43. 75% , -5. 56% , and stress peak of screw-rod fixation system were reduced by 47. 81% , 60. 90% , 48. 45% , 39. 14% , respectively. Under two internal fixation modes, stresses of C2 endplate and cage were mainly distributed on the compressive side during the motion, and both the screw-bone interface and the caudal side of screw subjected to large loading. Conclusions Two internal fixation methods could provide similar stability. However, the stress concentration of screw-rod system was more obvious and the possibility of screw loosening and fracture was greater under Cage+ C1LMS+C2PS fixation.