Bone density measurements are currently being performed throughout the world in the diagnosis and management of osteoporosis as well as in research into this major health problem. However, it is not clear to what extent bone mineral density (BMD) values determined by dual-photon absorptiometry at one center can be applied to another. This is particularly relevant for the quantitative comparison of results from studies carried out in different laboratories. Furthermore, many centers now acquiring densitometers may not have the resources to determine their own normal range, relying instead on a "normal" range provided by the manufacturer. The question of the comparability of BMD data obtained in different centers was examined by comparing the normal range for the lumbar spine and proximal femur in 203 normal white Australian women and 892 normal white U.S. women, obtained using the same model densitometer. The two populations were compared according to decade. From superimposition of the Australian individual values on the North American normal ranges, only minor differences between the two populations were seen at any of the sites measured at any decade. None of these minor differences were statistically significant. This study shows a close similarity between BMD values in both the proximal femur and lumbar spine in normal white women in Australia and North America, provided the same model densitometer is used. Thus data obtained from different centers in populations with similar ethnic composition may be compared directly. These findings provide for the first time a sound basis for the quantitative comparison of the at times conflicting studies carried out in widely differing settings around the world.