FThe purpose of this study is to clarify the destructive processes of periodontal tissue, especially of alveolar bone, induced by food impaction. Streptozotocin-induced diabetic rat was used, gutter percha point (GP) was inserted into interdental gingival col space between upper M1 and M2, and the change of body weight, histopathological findings and 3D SEM structure of alveolar bone were compared between diabetic (Group DM) and control (Group N) animals (Experiment 1). Large amounts of bacterial deposits and sequester were observed at the alveolar crest, and subsequently experiment 2 was carried out to explore the relationship between bacterial infection and sequester formation. The combination of mechanical compression and daily twice cleaning with 3% oxydol and Periocline were carried out. As a result of experiment 1, ulceration with partial exposure of alveolar crest, inflammatory infiltrates and hyaline degeneration occurred in the col, and bone resorption was scarcely observed at 1d in Group N. Bone resorption of alveolar bone progressed at 3d and 5d, decreasing the height of alveolar crest and the width of alveolar bone. Reepithelization of ulcer surface was observed at 7d and 14d, and concurrently bone resorption regressed and new bone formation suggested reparative changes. SEM observation confirmed these changes of alveolar bone. On the contrary, massive bacterial deposits were observed, and bone resorption was scanty in the upper region and slight in the middle to lower region of alveolar bone at 1 day of Group DM. Massive bacterial deposits with partial exposure of alveolar bone were observed at 3d and 5d, and sequester was isolated owing to intense undermining bone resorption. Regeneration of epithelium was seen beneath the sequester at 14d, showing the phenomenon of foreign body exclusion. The light microscopic changes of Group DM were consistent with SEM findings. The total number of osteoclasts was fewer at 1d and larger at 3d and 5d in Group DM. Osteoclasts at the upper region of alveolar bone increased from 1d to 5d in Group N, although the number of osteoclasts did not show a significant change in Group DM. At the middle to lower portion of alveolar bone, the number of osteoclast was significantly fewer at 1d, larger at 3d and 5d in Group DM. In experiment 2, the col with ulceration and necrosis did not show bacterial deposits or sequester, suggesting the intimate relation between bacterial infection and sequestration. The present study suggests that food impaction and subsequent mechanical compression to the col could induce various damages of periodontal tissue, including sequester formation, and elaborate plaque control at the initial stage of food impaction could prevent sequester formation and protect the height of alveolar bone.