Previous studies have indicated a relationship between bone mineral density and the incidence of breast cancer in middle-aged and elderly women, with women with higher BMD being at significant increased risk. We investigated whether there was such a relationship in younger women who were perimenopausal or in their early postmenopausal years. As part of a population-screening program for osteoporosis, 5,119 women aged between 45 and 54 years were scanned between 1990-1994 at the Osteoporosis Research Unit. In 1997-2001, 3,884 returned for follow-up scans and questionnaires, and 3,144 returned a postal questionnaire in 2002. All cases of incident breast cancer were noted. One hundred sixty-six women indicated that they had suffered from breast cancer, of which 87 were incident cases (59 had prevalent breast cancer at baseline and 20 had benign or unconfirmed diagnosis and were excluded because of the use of agents that may interfere with BMD, e.g., tamoxifen). We compared therefore the incident breast cancer group (BC group; n=87) with a control group (C group; n=3,013). There were no significant differences using a t-test between the BC group and C group for baseline DXA of the spine or femoral neck. Further changes in BMD over a mean period of 6.9 years demonstrated no significant hazard ratio for the lumbar spine or femoral neck. No relationship was seen between the bone turnover markers pyridinoline/creatinine or deoxypyridinoline/creatinine assessed at their second study visit and incidence of breast cancer. In conclusion, in perimenopausal or early postmenopausal women there is no relationship between the incidence of breast cancer and BMD, change in BMD or bone turnover.