BACKGROUND: Due to their known effects on bone metabolism, vitamin D and related compounds have been proposed for the prevention of osteoporosis and fractures. OBJECTIVES: To determine the effects of supplementation with Vitamin D or a Vitamin D analogue in the prevention of fractures of the axial and appendicular skeleton in elderly men or women with involutional or post-menopausal osteoporosis. SEARCH STRATEGY: We searched MEDLINE, EMBASE, CINAHL, LILACS, CABNAR, BIOSIS, HEALTHSTAR, Current Contents, The Cochrane Database of Systematic Reviews, the Cochrane Musculoskeletal Injuries Group trials register, and bibliographies of identified trials and reviews. Date of the most recent search: September 2000. SELECTION CRITERIA: Any randomised or quasi-randomised trial which compared vitamin D or a vitamin D analogue, either alone or in combination with calcium supplementation, with a placebo, no intervention, or the administration of calcium supplements, with eligible fracture outcomes, in elderly men or women with involutional or post-menopausal osteoporosis. DATA COLLECTION AND ANALYSIS: Two reviewers independently assessed trial quality, by use of a nine item scale, and extracted data. Additional information was sought from trialists. Where possible the data were pooled. Pooling of data, where it was admissible, used pooled relative risk and fixed effects model. MAIN RESULTS: Almost all estimates of treatment effects are based on single studies. Administration of vitamin D3 alone without calcium co-supplementation was not associated with any reduction in incidence of hip fracture (relative risk (RR) 1.20, 95% confidence interval (CI) 0.83, 1.75) or other non-vertebral fracture. Administration of vitamin D3 with calcium co-supplementation to frail elderly people in sheltered accommodation was associated with a reduction in incidence of hip fracture (RR 0.74, 95% CI 0.60, 0.91). In healthy younger, ambulant participants the effect on hip fracture is unknown (RR 0.36, 95% CI 0.01, 8.78), although there appears to be a significant overall effect on non-vertebral fracture incidence in this group ( RR 0.46, 95% CI 0.23,0.90). Calcitriol (1,25 dihdyroxy vitamin D) was effective in reducing the incidence of vertebral deformity (RR 0.49, 95% CI 0.25, 0.95). Calcitriol was more effective than calcium in reducing the frequency of new vertebral deformities during the third year of treatment (RR 0.28, 95% CI 0.15, 0.52). 1-alpha-hydroxy vitamin D was effective in reducing the incidence of non-vertebral fractures in a single small study of elderly people whose mobility was impaired by neurological disease (RR 0.12, 95% CI 0.02, 0.95). No statistically significant effects were found for other comparisons of vitamin D or its analogues against each other, with and without calcium supplementation. REVIEWER'S CONCLUSIONS: Uncertainty remains about the efficacy of regimens which include vitamin D or its analogues in fracture prevention. Particularly if co-supplementation of calcium is required, significant cost differences are likely to exist between regimens. Further large randomised trials are currently being conducted to clarify the effectiveness of community fracture prevention programmes employing vitamin D supplementation.