This large-scale postmarketing surveillance of raloxifene (60 mg/day) was conducted to assess the safety and effectiveness of raloxifene for long-term use in postmenopausal Japanese women with osteoporosis. The baseline examination included 6,967 women (mean age, 70.4 years). Participants completed observation after 6, 12, 24, and 36 months of therapy. Adverse drug reactions (ADR) were reported in 776 participants (11.14 %), with a total of 87 serious ADR cases occurring in 76 participants (1.09 %). The most frequently reported ADRs were edema peripheral (45/6,967, 0.65 %) and venous thromboembolism (11/6,967, 0.16 %). Of the 6,967 participants, 2,784 were included in the effectiveness analysis. Lumbar spine bone mineral density (BMD) increased significantly (p < 0.001, paired t test) compared with baseline at 6, 12, 24, and 36 months (2.51 %, 2.85 %, 4.76 %, and 3.51 %, respectively). Significant decreases in serum and urinary cross-linked amino-terminal telopeptide of type I collagen (NTX) and urinary deoxypyridinoline levels from baseline were observed at 3 months, followed by a significant decrease of serum bone alkaline phosphatase at 6 months [p < 0.001 for all comparisons except serum NTX (p = 0.011), Wilcoxon signed-rank test]. Early reductions in the biochemical markers of bone turnover (BTM) observed at 3 months with raloxifene treatment correlated negatively with subsequent increases in lumbar spine BMD at 1 year (r = -0.347, p = 0.008). The incidence of any new clinical fractures within 3 years was 1.18 % (82/6,967 participants). In summary, no new signals in safety were observed in the daily use of raloxifene. Moreover, the effectiveness profile of raloxifene was confirmed in practical use by this large-scale, long-term, postmarketing surveillance.