OBJECTIVE We evaluated the biomechanical effects of 4 instrumented configurations after induced atlantoaxial rotatory subluxation: transarticular screw fixation (T/A) and polyaxial C1 lateral mass and C2 pedicle screw and rod fixation (LC1-PC2) for atlantoaxial arthrodesis with unilateral and bilateral instrumentation. METHODS Three-dimensional intervertebral motion was tracked stereophotogrammetrically while 14 human cadaveric spine specimens underwent nonconstraining pure moment loading. Nondestructive loads were applied quasi-statistically in 0.25-Nm increments to a maximum load of 1.5 Nm during flexion-extension, right and left axial rotation, and right and left lateral bending. Hyperrotation injuries were created using torsional loads applied during left axial rotation until visible failure occurred. RESULTS In the normal condition, the values for angular range of motion, lax zone (zone of ligamentous laxity), and stiff zone (zone of ligamentous stretching) were similar in both groups in all directions of loading, with no significant differences (P > 0.05) between groups at C0-C1 or C1-C2. Both instrumentation systems (bilateral configurations) substantially stabilized angular motion at C1-C2 (P < 0.05) during all loading modes for the T/A group, and during all but right lateral bending (P = 0.072) for the LC1-PC2 group. The mean failure load for both intact and instrumented specimens was slightly greater, but not significant for the LC1-PC2 group compared with the T/A group (P > 0.14). CONCLUSION Both methods fixated atlantoaxial subluxation equally well. Compared with unilateral instrumentation, a bilateral configuration with the LC1-PC2 technique significantly increased stability during extension (P < 0.05). During axial rotation, bilateral T/A screws significantly increased stability compared with unilateral fixation (P < 0.02).