Background Bone is the most common site of metastatic disease associated with breast cancer (BC). Bisphosphonates inhibit osteoclast-mediated bone resorption, and novel targeted therapies such as denosumab, inhibit key pathways in the vicious cycle of bone metastases. Objectives To assess the effect of bisphosphonates on skeletal-related events (SREs), bone pain, quality of life (QoL), recurrence and survival in women with breast cancer with bone metastases (BCBM), advanced breast cancer (ABC) without clinical evidence of bone metastases and early breast cancer (EBC). To assess the effect of denosumab on SREs, bone pain and (QoL) in women with (BCBM). Search methods We searched the Specialised Register maintained by the Cochrane Breast Cancer Group (CBCGSR), MEDLINE, EMBASE and the WHO International Cancer Trials Registry Platform (WHO ICTRP) on 30 April 2011. We conducted additional handsearching of journals and proceedings of key meetings. Selection criteria We included randomised controlled trials (RCTs) comparing: (a) bisphosphonates and control, or different bisphosphonates in women with BCBM; (b) denosumab and bisphosphonates in women with BCBM; (c) bisphosphonates and control in women with ABC; (d) bisphosphonates and control in women with EBC; and (e) early versus delayed bisphosphonate treatment in women with EBC. Data collection and analysis Two review authors (MW and NP) independently assessed the trials and extracted the data. We collected toxicity information from the trials. Main results We included thirty-four RCTs. In nine studies (2806 patients with BCBM), comparing bisphosphonates with placebo or no bisphosphonates, bisphosphonates reduced the SRE risk by 15% (risk ratio (RR) 0.85; 95% confidence interval (CI) 0.77 to 0.94; P = 0.001). This benefit was most certain with intravenous (i.v.) zoledronic acid (4 mg) (RR 0.59; 95% CI 0.42 to 0.82); i.v. pamidronate (90 mg) (RR 0.77; 95% CI 0.69 to 0.87); and i.v. ibandronate (RR 0.80; 95% CI 0.67 to 0.96). A direct comparison of i.v. zoledronic acid and i.v. pamidronate confirmed at least equivalent efficacy in a single large study. In three studies (3405 patients with BCBM), compared with bisphosphonates, subcutaneous (s.c.) denosumab was more effective in reducing the risk of SREs (RR 0.78; 95% CI 0.72 to 0.85; P < 0.00001). Bisphosphonates reduced the SRE rate in 12 studies (median reduction 28%, range 14% to 48%), with statistically significant reductions reported in 10 studies. Women with BCBM treated with bisphosphonates showed significant delays in the median time to SREs. Compared with placebo or no bisphosphonates, treatment with bisphosphonates significantly improved bone pain in six out of eleven studies. Improvements in global QoL with bisphosphonates compared to placebo were reported in two out of five studies (both ibandronate studies). Treatment with bisphosphonates did not appear to affect survival in women with BCBM. Compared to i.v. zoledronic acid, denosumab also significantly reduced the SRE rate, delayed the time to SREs and prolonged the time in developing pain for patients with no or mild pain at baseline; but there was no difference in survival between patients treated with denosumab and zoledronic acid. Bisphosphonates in women with ABC without clinically evident bone metastases did not reduce the incidence of bone metastases, or improve survival in three studies (320 patients). In seven studies (7847 patients with EBC), currently there is no evidence supporting bisphosphonates in reducing the incidence of bone metastases compared to no bisphosphonates (RR 0.94; 95% CI 0.82 to 1.07; P = 0.36). In three studies (2190 patients with EBC), early bisphosphonate treatment also did not significantly reduce the incidence of bone metastases compared to delayed bisphosphonate treatment (RR 0.73; 95% CI 0.40 to 1.33; P = 0.31). Currently, there is insufficient evidence to make a conclusion about the role of adjuvant bisphosphonates in reducing visceral metastases, locoregional recurrence and total recurrence, or improving survival. There was strong heterogeneity in EBC studies examining the outcomes of total recurrence and survival. Reported toxicity was generally mild. Renal toxicity and osteonecrosis of the jaw (ONJ) have been identified as potential problems with bisphosphonate use. ONJ was reported at similar rates for patients on denosumab compared to zoledronic acid. This highlighted a need for maintaining good oral care, prior to and during treatment, for patients who received long-term bone agents. Authors' conclusions In women with clinically evident BCBM, bisphosphonates (oral and i.v.) and denosumab (s.c.) reduced the risk of developing SREs, as well as delaying the time to SREs. Some bisphosphonates may also reduce bone pain and may improve QoL. The optimal timing and duration of treatment for patients with BCBM remains uncertain. There is currently insufficient evidence to support the routine use of bisphosphonates as adjuvant treatment for patients with EBC. However, a number of large clinical trials investigating bisphosphonates in EBC have completed accrual and are awaiting results.