Questionnaires covering health and the use of complementary, alternative and conventional health services were mailed to a random sample of 14 868 adults aged 18-64 years living in four counties of England in 1997. The present study examined the use of osteopathy/chiropractic among the 15% (n = 1377) of respondents reporting back pain. Osteopaths/chiropractors were seen by 13.4% (n = 184) of respondents with back pain during the past 3 months compared with 9.8% (n = 135) who consulted physiotherapists. The presence of back pain and non-manual social class were the strongest predictors of consultation with both types of practitioner. Women, older respondents, non-smokers and those who exercised for 30 minutes at least once a week were more likely to use osteopathy/chiropractic. The only other significant predictor of physiotherapy use was desire for more physical exercise. While those reporting back pain had Short-Form 36 (SF-36) scores suggesting very significant levels ofdisability, respondents with back pain who consulted osteopaths/chiropractors reported better health in all dimensions of the SF-36 than those using physiotherapy services. Although they reported worse pain scores than people not consulting any practitioners, their mental health, physical functioning, energy and health perception were better. It is impossible to disentangle cause and effect in this cross-sectional study, but the data suggest that people who can afford to pay are more likely to choose osteopath/chiropractor treatments than physiotherapy. The possibility that osteopath/chiropractor treatment has a generalised positive effect on health, allowing people with back pain to function better than those not receiving such treatment, warrants further investigation.