Objective: To perform a meta-analysis exclusively including head-to-head randomized controlled trials (RCTs) that directly compared the fracture prevention efficacy and the safety between alendronate (Aln) and raloxifene (Rlx) for postmenopausal women with osteoporosis,and comprehensively identify the potential influential factors.Data sources: Electronic databases (Pubmed,Medline,Clinical Trial Registry,EMBASE and the Cochrane Data Base of Systematic Reviews,the Cochrane Central Register of Controlled Trials) were searched without limit up to September 2012 for relevant articles.Review methods: Only RCTs containing both Aln and Rlx for treating postmenopausal women with osteoporosis,and assessing at least one of the following outcomes,fracture incidence,bone mineral density (BMD),bone turnover markers (BTM),serum lipid data,and safety,were identified and considered for inclusion.We evaluated the methodological quality,and performed analysis using Review Manager 5.1 and STATA 11.0.Results: Seven RCTs involving 4054 women met our inclusion criteria.The evidence of moderate quality suggested no superiority of Aln over Rlx in reducing the risk of both vertebral fractures [risk ratio: 1.30 (0.66-2.54)] and non-vertebral fractures [0.95 (0.54,1.68)].Also the evidence of moderate to high quality showed Aln was more effective in increasing BMD and inhibiting BTM than Rlx within a follow-up of 12-24 months.Aln reduced the risk of vasomotor by 57%[2.33 (1.21-4.49)] but increased the risk of diarrhea by 133%[0.47 (0.27-0.81)] compared to Rlx,which was supported by the evidence of moderate to high quality.Rlx has additional benefits in modulating the serum lipid concentration [total cholesterol: 2.36 (1.18-3.53),triglycerides:-5.60 (-15.87-4.67),low density lipoprotein cholesterol: 8.16 (7.23-9.08),high density lipoprotein cholesterol:-5.22 (-6.12-4.33)].Our subgroup analysis further indicated the difference between Aln and Rlx in fracture reduction and BTM inhibition were not materially altered by the administration pattern,the age,the methodological quality,the sample size,or the industrial funding.The weekly strategy of Aln would further reduce the upper GI disorders [weekly: 1.34 (1.04-1.72),daily: 0.92 (0.49-1.72)] and might gain more bone mass increment at lumbar spine (0.36) compared to its daily treatment.Conclusion: There was equivalent risk reduction between Aln and Rlx in the vertebral and non-vertebral fracture.In addition,the age did not obviously influence their relative anti-fracture efficacy.During clinical decision making,the side-effects and additional benefits associated with both drugs,which affected the patients adherence,should also be taken into account.