Distraction osteogenesis has proven to be an effective technique for the correction of mandibular deficiencies. However, problems have been encountered in achieving a final, idealized form of the mandible when using distraction devices capable of moving the bone segments in only one dimension (uniplanar). Specifically, occlusal irregularities and deficiencies in lower facial contour have been seen following uniplanar distraction. To address these problems, a distraction device capable of independent movements in three planes (multiplanar) was developed. Previously reported studies in a canine model have demonstrated that this device can successfully distract the mandible along both the sagittal axis (anteroposterior or z-axis) and the vertical axis (superoinferior or y-axis). This study examines the ability of the multiplanar device to distract along the sagittal and horizontal axes (mediolateral or x-axis). A total of 12 dogs were included in the study. All animals underwent unilateral or bilateral mandibular distraction using an external multiplanar device. After a latency period of 5 days, primary distraction along the anteroposterior axis at a rate of 1 mm/day for 10 days (10 mm total) was performed. During the following 10 days, along with an additional 11 mm to 20 mm of anteroposterior axis distraction, concomitant secondary distraction was performed along the horizontal (mediolateral) axis at a rate of 5 degrees/day (50 degrees total). Cephalometric radiographs were obtained preoperatively and at the conclusion of both anteroposterior and combined anteroposterior-mediolateral distraction. Computed tomography (CT) scans were obtained preoperatively and at the end of consolidation (28 days), after which all animals were sacrificed and the dry skulls examined. In all animals, distraction along the mediolateral or x-axis was found to change the anteroposterior projection of the mandible. Varus angulation of the device with respect to the midline of the mandible caused compression of the distracted segments and reduced the anteroposterior thrust of the mandible. In contrast, valgus positioning of the device, with respect to the midline of the mandible, created the opposite effect, increasing the distracted length in the anteroposterior direction. The bone (mandibular) segments being distracted assumed the orientation of the device only for valgus positioning of the device (producing a decrease in the bigonial distance). Conversely, there was no effect from the mediolateral angulation on the distracted segments during varus positioning of the device. A possible explanation for this finding may be a greater resistance to an increase in the bigonial distance (varus positioning of the device) posed by obstruction of lateral movement of the condyle. This stands in contrast to a decrease in the bigonial distance observed following valgus positioning of the device. These findings confirm the clinical impression that distraction along the anteroposterior or sagittal axis remains the critical or keystone therapeutic maneuver in distraction of the mandible. Mediolateral or horizontal axis distraction is best used only as a supplementary movement; in essence, it only affects the anteroposterior dimension with little impact on clinically relevant changes to the bigonial distance.