Poor compliance may be detrimental to the effectiveness of a screening programme if those at greatest risk of the disease do not attend. Therefore we undertook a study to test whether non-attenders to a screening programme for low bone density, a risk factor for osteoporosis, were at a differential risk of low bone density compared with attenders. Seven hundred and eighty-nine women aged 45-49 years living within 32 km of Aberdeen were selected at random from the Community Health Index and invited to attend for screening for low bone density as a risk of factor for osteoporosis. Attenders and non-attenders were surveyed regarding their risk factors for osteoporosis. Non-attenders were significantly heavier than attenders. In addition, 6 non-attenders who subsequently chose to attend had significantly higher body weight and bone mineral density, at Ward's triangle, than initial attenders. Non-attenders to a screening service for bone density may be at lower risk of developing osteoporosis. Non-attendance, therefore, would not be detrimental to the cost-effectiveness of a screening service for bone density. However, this study indicates there is a potential for response bias in studies of bone density and osteoporosis.