Background: To evaluate the association between type 2 diabetes and periodontal disease severity using the rate of alveolar bone loss (ABL) and high-sensitivity C-reactive protein (hs-CRP) value as indices. Methods: In this cross-sectional study of 372 subjects (mean age ± SD, 53.2 ± 11.8 years) from a Japanese hospital, we measured ABL and number of teeth on either panoramic radiographs or intraoral dental radiographs of all teeth. A trial classification of periodontal disease was defined as follows: I (ABL < 25%, clinically mild); II (ABL ≥ 25%, <35%, moderate); III (ABL ≥ 35%, severe); and A (hs-CRP < 440 ng/ml, mild inflammation); B (hs-CRP ≥440 ng/ml, <1,020 ng/ml, moderate inflammation); and C (hs-CRP ≥ 1,020 ng/ml, severe inflammation). Periodontal disease severity was classified into nine groups by combining ABL and hs-CRP. Results: Forty-eight subjects had type 2 diabetes; 324 did not. Univariate analysis showed that type 2 diabetes was significantly associated with age, sex, body mass index, number of teeth, ABL, hs-CRP, and periodontal disease severity. Multivariate analysis showed significant associations between type 2 diabetes and the groups with high severity of periodontal disease. In receiver operating characteristic (ROC) curve analysis predicting the presence of diabetes, area under the ROC curve was 0.762 (95%CI = 0.688 - 0.835) for ABL, and 0.709 (95%CI = 0.635 - 0.784) for hs-CRP, which was significant. Conclusions: This study showed that diabetes could be associated with a periodontal disease severity classification using the combination of ABL and hs-CRP.