Recently, various types of semi-rigid pedicle screw fixation systems have been developed for the surgical treatment of the lumbar spine. They were introduced to address the adverse issues commonly found in traditional rigid spinal fusion--abnormally large motion at the adjacent level and subsequent degeneration. The semi-rigid system uses more compliant materials (nitinol or polymers) and/or changes in rod design (coiled or twisted rods) as compared to the conventional rigid straight rods made of Ti alloys (E = 114 GPa, υ = 0.32). However, biomechanical studies on the semi-rigid pedicle screw systems were usually limited to linear modeling of the implant and anatomic elements, which may not be capable of reflecting realistic post-operative motions of the spine. In this study, we evaluated the effects of nonlinearity in materials used for semi-rigid pedicle screw fixation systems to evaluate the changes in biomechanical behaviors using finite element analysis. Changes in range of motion (ROM) and center of rotation (COR) were assessed at the operated and adjacent levels. Actual load-displacement results of the semi-rigid rod from mechanical test were carried out to reflect the nonlinearity of the implant. In addition, nonlinear material properties of various spinal ligaments studies were used for the finite element modeling. The post-operative models were constructed by modifying the previously validated intact model of the L1-S1 spine. Eight different post-operative models were made to address the effects of nonlinearity-with a traditional stiffness modulus rod (with linear ligaments, case 1; with nonlinear ligaments, case 5), with a rigid rod (with linear ligaments, case 2; with nonlinear ligaments, case 6), with a soft rod (with linear ligaments, case 3; with nonlinear ligaments, case 7), and with a nonlinear rod (with linear ligaments, case 4; with nonlinear ligaments, case 8). To simulate the load on the lumbar spine in a neutral posture, follower load (400 N) was applied and then the hybrid loading condition was applied to measure the ROM and COR in the sagittal plane. The more the nonlinearity was included in the model the closer the motion behavior of the device was to that of the intact spine. Furthermore, our results showed that the nonlinearity of the semi-rigid rod was a more sensitive factor than the nonlinearity of the spinal ligaments on biomechanical behavior of the lumbar spine after surgery. Therefore, for better understanding of the surgical effectiveness of the spinal device, more realistic material properties such as nonlinearity of the device and anatomic elements should be considered. In particular, the nonlinear properties of the semi-rigid rod were considered more than the nonlinearity of spinal ligaments.