Dynamic stabilization devices have been introduced to clinics as an alternative to rigid fixation. The stiffness of these devices varies widely, whereas the optimal stiffness, achieving a predefined stabilization of the spine, is unknown. This study was focused on the determination of stiffness values for posterior stabilization devices achieving a flexible, semi-flexible or rigid connection between two vertebrae. An extensively validated finite element model of a lumbar spinal segment L4-5 with an implanted posterior fixation device was used in this study. The model was exposed to pure moments of 7.5 and 20Nm around the three principal anatomical directions, simulating flexion, extension, lateral bending and axial rotation. In parametrical studies, the influence of the axial and bending fixator stiffness on the spinal range of motion was investigated. In order to examine the validity of the computed results, an in-vitro study was carried out. In this, the influence of two posterior stabilization devices (DSS and rigidly internal fixator) on the segmental stabilization was investigated. The finite element (FE)-model predicted that each load direction caused a pairing of stiffness relations between axial and bending stiffness. In flexion and extension, however, the bending stiffness had a neglectable effect on the segmental stabilization, compared to the axial stiffness. In contrast, lateral bending and axial rotation were influenced by both stiffness parameters. Except in axial rotation, the model predictions were in a good agreement with the determined in-vitro data. In axial rotation, the FE-model predicted a stiffer segmental behavior than it was determined in the in-vitro study. It is usually expected that high stiffness values are required for a posterior stabilization device to stiffen a spinal segment. We found that already small stiffness values were sufficient to cause a stiffening. Using these data, it may possible to develop implants for certain clinical indications.