The authors conducted a systematic review of published data on the association between diabetes mellitus and fracture. The authors searched MEDLINE through June 2006 and examined the reference lists of pertinent articles (limited to studies in humans). Summary relative risks and 95% confidence intervals were calculated with a random-effects model. The 16 eligible studies (two case-control studies and 14 cohort studies) included 836,941 participants and 139,531 incident cases of fracture. Type 2 diabetes was associated with an increased risk of hip fracture in both men (summary relative risk (RR) = 2.8, 95% confidence interval (CI): 1.2, 6.6) and women (summary RR = 2.1, 95% CI: 1.6, 2.7). Results were consistent between studies of men and women and between studies conducted in the United States and Europe. The association between type of diabetes and hip fracture incidence was stronger for type 1 diabetes (summary RR = 6.3, 95% CI: 2.6, 15.1) than for type 2 diabetes (summary RR = 1.7, 95% CI: 1.3, 2.2). Type 2 diabetes was weakly associated with fractures at other sites, and most effect estimates were not statistically significant. These findings strongly support an association between both type 1 and type 2 diabetes and increased risk of hip fracture in men and women.