Using a rabbit model, we evaluated the role of the periosteum in bone induction using hydroxyapatite ceramic pellets, some of which had been coated with bone morphogenetic protein. Eighteen rabbits were divided into two groups, a control group that received pellets soaked in phosphate-buffered saline alone and another group for which the pellets had been supplemented with bone morphogenetic protein. After making a skin incision on the head of each rabbit, two caudally pedicled periosteal flaps measuring 1 cm in width and 2.5 cm in length were elevated. These flaps were wrapped around hydroxyapatite pellets manufactured from limestone and fixed with sutures. After implantation, both groups of rabbits were returned to their cages, maintained for 3, 6, or 9 weeks [every group consisted of 3 rabbits (6 pellets)], and then sacrificed. In this study, the extent of bone induction, which was measured with a dual x-ray densitometer, that resulted from covering the hydroxyapatite pellet with periosteum alone (control group) was minimal. On the contrary, since osteogenesis from the periosteum toward the pores of the pellet was observed in the bone morphogenetic protein group much more than in the control group, the usefulness of our technique was confirmed. However, active osteogenesis was observed with subperiosteal implantation of the hydroxyapatite-bone morphogenetic protein complex in the bone morphogenetic protein group, but the osteogenesis observed was not distributed over the entire pellet.