Bone‐specific drugs (BSDs) increase the risk of osteonecrosis of the jaw (ONJ), but whether they increase the risk of osteonecrosis at other sites is not known. Two studies, a cohort study and a case–control study, were conducted using registry data on everyone who was residing in Sweden on December 31, 2005, and who was 50 years of age or older at the time (n = 3,523,912). In the cohort study, individuals prescribed a BSD during the period 2006–2017 (n = 217,387) were 1:​1 matched with nonusers on birth year, sex, hip fracture status, and Swedish or foreign origin. In the case–control study, individuals diagnosed with osteonecrosis during 2006–2017 (n = 12,614) were 1:​1 matched with individuals without a diagnosis of osteonecrosis on birth year, sex, and Swedish or foreign background. In the cohort study, osteonecrosis was diagnosed in 983 BSD users and 214 nonusers (adjusted hazard ratio [aHR] 4.02;​ 95% CI, 3.32–4.87), during a mean treatment time of 2.8 years. A similar association was observed in a subcohort where all individuals diagnosed with cancer (HR 4.82;​ 95% CI, 2.52–9.22). The greatest difference in incidence between BSD users and nonusers was observed in patients with a femoral neck fracture that was not treated with total hip arthroplasty or hemiarthroplasty (incidence rate difference, 77.8 cases per 10,000 person‐years, p < .05). The risk of osteonecrosis was higher in users of denosumab (HR 1.93;​ 95% CI, 1.33–2.79) and users of zoledronic acid (HR 1.95;​ 95% CI, 1.31–2.91) than in users of other BSDs. The increased risk of osteonecrosis decreased after the end of therapy (p < .001 for time trend). The results were confirmed in the case–control study. In summary, use of BSDs, especially more potent BSDs, is associated with increased risk of osteonecrosis of sites other than the jaw. This increased risk decreases after the final dose of BSD.