BACKGROUND: Fractures of the hip are an important cause of later ill health and mortality in elderly people. People with hip fractures are often malnourished at the time of fracture, and have poor food intake in hospital. OBJECTIVES: This review assesses the effects of nutritional interventions in elderly people recovering from hip fracture. SEARCH STRATEGY: We searched the Cochrane Musculoskeletal Injuries Group specialised register, the Cochrane Central Register of Controlled Trials (The Cochrane Library issue 2, 2003), MEDLINE (1966 to July 2003), Nutrition Abstracts and Reviews (1984 to March 2003), EMBASE (1980 to week 29 2003), BIOSIS (1985 to July 2003), CINAHL (1982 to July 2003), HEALTHSTAR (1975 to March 2002), the National Research Register and reference lists. We contacted investigators, and handsearched the American Journal of Clinical Nutrition, Proceedings of the Nutrition Society, Clinical Nutrition and the Journal of Parenteral and Enteral Nutrition. SELECTION CRITERIA: Randomised and quasi-randomised trials of nutritional interventions of mainly older patients (aged over 65 years) with hip fracture. DATA COLLECTION AND ANALYSIS: Trial allocation to included, excluded and awaiting assessment categories, was by consensus. Both reviewers independently extracted data and assessed trial quality. Additional information was sought from all trialists. Pooling of data for primary outcomes and select exploratory analyses were undertaken. MAIN RESULTS: Seventeen randomised trials involving 1266 participants were included. Overall the quality of trials was poor; specifically in terms of allocation concealment, assessor blinding and intention to treat analysis. This, and the limited availability of outcome data, mean that the following results must be interpreted with caution. Oral multinutrient feeds (providing non-protein energy, protein, some vitamins and minerals), evaluated by seven trials, may reduce unfavourable outcome (death or complications) (14/66 versus 26/73; relative risk 0.52, 95% confidence interval 0.32 to 0.84), but did not demonstrate an effect on mortality (12/91 versus 14/97; relative risk 0.85, 95% confidence interval 0.42 to 1.70). Four trials, examining nasogastric multinutrient feeding, showed no evidence for an effect on mortality (relative risk 0.99, 95% confidence interval 0.50 to 1.97), but the studies were heterogeneous regarding case-mix. Insufficient information was provided to evaluate unfavourable outcome. The effect of protein in an oral feed, tested in three trials, showed no evidence for an effect on mortality (relative risk 1.38, 95% confidence interval 0.82 to 2.34). It may have reduced the number of long term complications and days spent in rehabilitation wards. Two trials, testing intravenous thiamin (vitamin B1) and other water soluble vitamins, or 1-alpha-hydroxycholecalciferol (an active form of vitamin D) respectively, produced no evidence of benefit for either vitamin supplement. REVIEWER'S CONCLUSIONS: The strongest evidence for the effectiveness of nutritional supplementation exists for oral protein and energy feeds, but the evidence is still very weak. Future trials are required which overcome the defects of the reviewed studies, particularly inadequate size, methodology and outcome assessment.