A decreased risk of breast cancer has been reported among patients given bisphosphonates. The present aims were to study potential associations between different antiosteoporosis drugs, including bisphosphonates, and the risk of breast cancer before and after start of treatment and to appraise possible dose-effect relationships. From national Danish registers, all female users of bisphosphonates aged ≥40 years and other drugs against osteoporosis between 1996 and 2006 were identified (n = 87,104). This cohort was compared with a control group, where each patient was matched on age with three nonexposed women from the general population (n = 261,322). Before start of most drugs against osteoporosis an increased risk of breast cancer was seen compared to controls (e.g., adjusted OR = 1.09, 95% CI 1.04-1.16 for alendronate). This excess risk was higher in younger women (e.g., OR = 4.48, 95% CI 2.98-6.75 for alendronate in women ≤50 years) and disappeared in women older than 70 years (e.g., OR = 0.95, 95% CI 0.88-1.01 for alendronate). In contrast, a decreased risk of breast cancer was seen after start of alendronate (HR = 0.53, 95% CI 0.38-0.73), etidronate (HR = 0.80, 95% CI 0.73-0.89), and raloxifene (HR = 0.53, 95% CI 0.38-0.73). No dose-response relationship was present for alendronate and etidronate, whereas a decreasing risk was seen with increasing daily dose of raloxifene. Bisphosphonate treatment in women was associated with a reduced risk of breast cancer. However, no causal relationship seemed to be present.