Abstract The WHO criteria for osteoporosis are based on bone mineral density (BMD) values in comparison to a reference population of healthy young adults. The aim of this study was to create BMD references for ethnic Swedish women, and to investigate whether the use of these T-score measurements influence the amount of Swedish postmenopausal patients that are diagnosed as having osteoporosis. A bone density reference was created by measuring a population-based sample of 335 randomly selected Swedish women aged 20–39 yr. BMD was measured by dual-energy X-ray absorptiometry (DXA) at the lumbar spine, proximal femur, and total body. These locally derived T-score values were subsequently used to diagnose a sample of 300 consecutive postmenopausal Swedish patients referred to the Uppsala Osteoporosis Unit. There was a slight age-dependent decrease in femoral neck BMD, whereas no age effect was seen at other sites such as total hip, lumbar spine, or total body. This suggests that the cohort represents the steady state BMD at the ages of expected peak bone mass in Swedish women. The correlation between BMD measures at different sites differed from r = 0.55 (lumbar spine BMD vs femoral neck BMD [FNBMD]) to r = 0.92 (total hip BMD vs FNBMD). Central DXA-generated T-scores were calculated from this cohort, and these were significantly higher (0.3–0.5 SD) as compared with manufacturers and NHANESIII reference populations. This indicates that young Swedish women have a higher peak bone mass than the subjects included in the reference populations currently used for clinical measurements. The T-score in total hip derived from the investigated cohort was subsequently used to diagnose 300 clinical patients (mean age 63 yr) referred for a DXA scan by their physicians. The use of this locally established and ethnic representative, T-score reference increased the prevalence of osteoporosis in femoral neck and total hip with 53–106%. A Swedish female BMD reference representing peak bone mass has been established and the normative data are presented. Notably, this cohort has considerably higher BMD as compared to the NHANESIII and manufacturer's reference populations. The use of the present T-score reference therefore causes approximately a 2-fold increase in the amount of Swedish postmenopausal women that fulfill the WHO criteria for osteoporosis. This demonstrates the problems with using T-score as diagnostic threshold for osteoporosis and is an argument for future strategies to obtain standardized densitometric cut-offs, for example, mg/cm 2.