Oral bisphosphonates (BPS) represent the first line of prevention and treatment for osteoporosis. However, several studies have reported osteonecrosis of the jaw (ONJ), also known as osteomyelitis of the jaw (OMJ), as a side effect of these drugs. Although absolute risk is suggested to be low, information to date on the relative risk or attributable risk has in fact been limited. Here, we estimated risk of oral BPs for OMJ in osteoporosis patients taking oral BPs compared with other osteoporosis drugs. Using an electronic medical records retrieval system and manual confirmation of OMJ, we conducted a retrospective cohort study of patients taking medications for osteoporosis at Kyoto University Hospital between November 2000 and October 2010. To evaluate relative risks of oral BPs for OMJ, logistic regression analysis was performed and odds ratios (ORs) and 95% confidence interval (CIs) were estimated. In addition, sensitivity analyses were performed according to hierarchical diagnosis. A total of 4129 patients (59.6%) were prescribed BPs while 2794 (40.3%) received other osteoporosis drugs. Absolute risk for OMJ was estimated to range from 0.46% to 0.99% (95% CIs: 0.25-0.66 to 0.69-1.2) among patients receiving oral BPs and 0.071% to 0.17% (95% CIs: 0-0.17 to 0.022-0.33) among patients receiving other osteoporosis drugs. The attributable risks of oral BPs for OMJ were estimated to range from 0.38% to 0.81% (95% CIs: 0.38-0.39 to 0.80-0.81). ORs (95% CIs) adjusted for confounding factors were 5.0 (1.9-12.9) to 6.0 (1.3-26.1) for confirmed cases only. In terms of absolute and attributable risks, the risk of oral BPs for OMJ is considered to be less than 1% in patients with osteoporosis. However, oral BPs may increase the risk of OMJ compared with patients treated with other osteoporosis medications and the risk of side effects should be kept in mind. (C) 2012 Elsevier Inc. All rights reserved.