BACKGROUND: Hip fracture in the elderly is usually the result of a simple fall. Hip protectors have been advocated as a means to reduce the risk of sustaining a fracture in a fall on the hip. OBJECTIVES: To determine if external hip protectors reduce the incidence of hip fractures in elderly persons following a fall. SEARCH STRATEGY: We searched the Cochrane Musculoskeletal Injuries Group trials register (April 2003), the Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 1, 2003), MEDLINE (1966 to April 2003), EMBASE (1988 to 2003 week 14), CINAHL (1982 to April 2003) and reference lists of relevant articles. Trialists were contacted, and ongoing trials identified in The National Research Register (http://www.update-software.com/national/ accessed 20/01/03) and Current Controlled Trials (http://controlled-trials.com/ accessed 20/01/03). SELECTION CRITERIA: All randomised or quasi-randomised controlled trials (RCTs) comparing the use of hip protectors with a control group. DATA COLLECTION AND ANALYSIS: Two reviewers independently assessed trial quality, using a 10-item scale, and extracted data. Additional information was sought from all trialists. Wherever appropriate and possible, the data are presented graphically. MAIN RESULTS: In this third update, 13 randomised controlled trials were included. One, which was a study of compliance lasting 12 weeks, contributed no fracture outcome data. Five studies involving 4316 participants were cluster randomised by care unit, nursing home or nursing home ward rather than by the individual. Individually, each of these studies reported a reduced incidence of hip fractures within those units allocated to receive the protectors. Because of the use of cluster randomisation pooling of results of these studies was not undertaken. Pooling of data from five individually randomised trials conducted in nursing/residential care settings (1426 participants) showed no significant reduction in hip fracture incidence (hip protectors 37/822, controls 40/604, RR 0.83, 95% CI 0.54 to 1.29). Two individually randomised trials of 966 community dwelling participants, reported no reduction in hip fracture incidence with the hip protectors (RR 1.11, 95% CI 0.65 to 1.90). No important adverse effects of the hip protectors were reported but compliance, particularly in the long term, was poor. REVIEWER'S CONCLUSIONS: There is no evidence of effectiveness of hip protectors from studies in which randomisation was by individual patient within an institution, or for those living in their own homes. Data from cluster randomised studies indicates that, for those living in institutional care with a high background incidence of hip fracture, a programme of providing hip protectors appears to reduce the incidence of hip fractures. Acceptability by users of the protectors remains a problem, due to discomfort and practicality. Cost effectiveness remains unclear.