The present study was performed to evaluate the effects of hydroxylapatite (HAP) on tissue regeneration in various types of furcation involvement. Upper premolar tooth sites of 9 beagle dogs (3-6 years old) were used. HAP particles were implanted into three types of furcation defects: Class III lesion (by Glickman) in 2nd premolar sites, artificially caused by intrafurcal suturing or spontaneously developed about 3 months after extraction of the 1st premolar; Class IV lesion in 4th premolar sites artificially produced 2 months before implantation; and through-and-through furcal bony defects in 3rd premolar sites at the time of HAP implantation. Macroscopic, radiographic and microscopical investigations of the postoperative status were carried out at intervals of 1, 2, 3 and 6 months. The appearance of the gingiva in bony defect sites was almost the same as that in the presurgical status 2 weeks after implantation, although gingival inflammation was persistent in Class III and IV lesions. Most particles in Class III and IV lesions exfoliated until 2 weeks after implantation, and the junctional space between recipient bone and particles could not be distinguished, probably as a result of incorporation of HAP and osseous tissue. Histological observation revealed that HAP particles were surrounded by new bone located at the top of the alveolar crest in all defects. However, there were no obvious signs of coronal bone formation or connective tissue attachment in Class III and IV lesions up to 3 months after the operation. At 6 months there was evidence of new bone formation over the presurgical crests in Class III and IV lesions. On the other hand, there were obvious signs of connective tissue attachment and bone formation in bony defect sites, probably induced by the tissue of the periodontal ligament remaining after the surgical procedure, and nonphysiological ankylosis between root and bone was frequently observed. There were notable findings, such as newly formed osteoid tissue intervening with the adhesive particles, calcified tissue intervening between HAP particles and root surface showing ankylosis, and peripheral osteoid formation in the HAP particles. It is postulated that HAP has no apparent role in induction of bone formation, although there is chemical affinity to calcified tissue, and it is effective in yielding a volume of bone-like tissue where osseous repair could be performed. However, HAP did not enhance regeneration of lost periodontal structures including connective tissue attachment.