Recently, distraction osteogenesis has been widely used in the craniomaxillofacial skeleton. There are varying views, however, about the need for a latency period for mandibular distraction, which has been reported from 0 to 14 days. We used a rabbit model to determine the ideal latency period for mandibular distraction osteogenesis. We performed a subperiosteal corticotomy of the bilateral mandible and then applied an external device. We used 0-, 2-, 5-, and 10-day latency periods. Distraction was done at a rate of 0.5 mm per day (0.25 mm every 12 h) to elongate by 5 mm after each latency period. Hematoxylin-eosin stain was used for histologic observation, and bone labelling was observed with a fluorescence microscope. In the 0-day latency, the lengthened region was filled with fibrous tissue. Cartilage formed like isles along the lengthened edge. In the 10-day latency, the lengthened region was almost completely filled with woven bone. In the 2- and 5-day latency, there were mineralization zones at the periphery of this fibrous interzone. The primary osteon remodelling zones were located at both ends of the zones of mineralization zones. In the 5-day latency, new bone formation was observed around the periosteum.