BACKGROUND: In the United Kingdom (UK), 9% of adults consult their doctor annually with back pain. The treatment recommendations are based on orthopaedic teaching, but the current management is causing increasing dissatisfaction. Many general practitioners (GPs) are confused about what constitutes effective advice. AIM: To review all randomized controlled trials of bed rest and of medical advice to stay active for acute back pain. METHOD: A systematic review based on a search of MEDLINE and EMBASE from 1966 to April 1996 with complete citation tracking for randomized controlled trials of bed rest or medical advice to stay active and continue ordinary daily activities. The inclusion criteria were: primary care setting, patients with low back pain of up to 3 months duration, and patient-centred outcomes (rate of recovery from the acute attack, relief of pain, restoration of function, satisfaction with treatment, days off work and return to work, development of chronic pain and disability, recurrent attacks, and further health care use). RESULTS: Ten trials of bed rest and eight trials of advice to stay active were identified. Consistent findings showed that bed rest is not an effective treatment for acute low back pain but may delay recovery. Advice to stay active and to continue ordinary activities results in a faster return to work, less chronic disability, and fewer recurrent problems. CONCLUSION: A simple but fundamental change from the traditional prescription of bed rest to positive advice about staying active could improve clinical outcomes and reduce the personal and social impact of back pain.