The optimum fixation method to achieve atlantoaxial fusion after odontoid resection is still subject to discussion. Isolated posterior surgical procedures for treatment of irreducible atlantoaxial kyphosis with spinal cord compression are associated with a high rate of morbidity and mortality. Transoral atlantoaxial plate fixation has been described by Harms as a fixation technique after odontoid resection. The purpose of this study was to compare biomechanically a new anterior atlantoaxial locking plate (AALP) with the Harms plate and the atlantoaxial transarticular screw fixation described by Magerl. Sixteen human cadaver craniocervical specimens were tested in flexion, extension, rotation, and bending with a nondestructive flexibility method using a nonconstrained testing apparatus. Five different groups were examined: (1) control group (intact) (n = 16), (2) unstable group (after dissection of the atlantoaxial ligaments and odontoidectomy) (n = 16), (3) Harms group (transoral atlantoaxial plate fixation) (n = 8), (4) AALP group (anterior atlantoaxial locking plate fixation) (n = 8), and (5) Magerl group (transarticular atlantoaxial screw fixation) (n = 16). Stiffness in any direction was significantly higher in the AALP specimens and those fixated with the Magerl method than in the Harms, control, or unstable specimens. The difference in stiffness between the AALP and Magerl reconstruction groups was not statistically significant. Experimentally, the AALP was significantly stiffer than the Harms plate. The AALP provided stiffness equal to transarticular screw fixation according to Magerl. A question that cannot be answered by this in vitro study concerns the level of rigidity required to obtain long-term stability and fusion by atlantoaxial fixation methods. However, it may be assumed that the more spinal motion is eliminated the greater the chance of definite spinal fusion. Therefore, it seems reasonable that the most reliable and rigid fixation method would be the method of choice.