Macroscopic, roentgenographic and histopathological studies were undertaken to evaluate the effectiveness of Hydroxyapatite Ceramics (HAP) and Polycrystal Alumina Ceramics (Al_2O_3) as a bone substitutes implanted in the extraction sockets where the periapical lesions were experimentally induced before extraction of teeth in Macaca fuscata fuscata and to compare the results with those for control sites which were not implanted. The results were as follows : 1. Macroscopic findings On the 14th postoperative day the postoperative reactive inflammation almost disappeared in both control and implantation groups. The narrowing of width of alveolar ridge and its lowering were observed after operation in the control group, but in both implantation groups the narrowing of width was scanty and the level of alveolar ridge was maintained high. 2. Roentgenographic findings The extraction sockets showed an increase of radio-opacity in process of time and the similar findings to those of the surrounding bone in the control group, but alveolar crest showed depression. In the implantation groups, HAP showed the granulated radioopaque image with large and small irregular in shapes, and Al_2O_3 showed smaller grobular granulated radio-opaque image than HAP. They existed at the level of ridges in extraction sockets. 3. Histopathological findings 1) On the 1st postoperative day of blood clotting stage the extraction sockets were filled with blood clots in the control group, whereas in the implantation groups they were filled with numerous HAP and Al_2O_3 granules, which exsisted at the high level of alveolar crest. On the 5th postoperative day the formation of new bone trabeculae in the extraction sockets floor were seen in the control group, but in the implantation groups it was seen on the 7th postoperative day. The amount of bone formation was greater in HAP implantation group than in Al_2O_3 implantation group. In both control and implantation groups the residual inflammatory lesions were observed in the extraction sockets floor with reduction in extent of inflammatory infiltration in process of time. 2) On the 14th postoperative day of the granulation tissue stage the extraction wounds were completely covered with epithelium in both control and implantation groups. In the extraction sockets of control group, the coronal about one-half layer consisted of granulation tissue, whereas about one-half of apical layer consisted of new bone trabecula. In the HAP group, greater part of layer consisted of granulation tissue with new bone trabeculae in about one-fifth of the apical layer, whereas in the Al_2O_3 implantation group the amount of regions occupied with new bone trabeculae was less. HAP and Al_2O_3 granules exsisted in large numbers within the granulation tissue and new bone trabeculae, reaching the portion directly underneath the epithelial layer. The inflammatory lesions in the extraction sockets were indistinct in both control and implantation groups. 3) On the 21th and 28th postoperative day of trabecula stage, the formation of new bone trabeculae proceeded in both control and implantation groups, but it was fairly delayed in the implantation group as compared with that of control group. In the implantation group, the amount of bone formation in HAP implantation group was greater than that in Al_2O_3 implantation group. HAP and Al_2O_3 granules were observed in large numbers within granulation tissue and new bone trabecula of extraction sockets, existing in the portion directly underneath the epithelial layer. No inflammatory cellular infiltration was observed around HAP and Al_2O_3 granules, and their cementing granulation tissue and new bone trabecula were dense. 4) On the 60th postoperative day of healing stage, the epithelia of extraction wounds were nearly normal in both control and implantation groups. The extraction sockets were filled with trabeculae which proceeded in