Background: Bone mineral density (BMD) is an important index in diagnosing osteoporosis and other metabolic bone diseases, predicting fractures, and monitoring treatment. Dual-energy X-ray absorptiometry (DXA) is most widely used technique for assessments of BMD and considered as the gold standard for the minumum exposure to radiation, low cost, high reproducibility, and ease of use. Author performed an experimental study to determine the relationship between the change in BMD measured by DXA and real change in mineral of human long bone. Materials and Methods: A pair of humeri and femora from one male cadaver was cut into specimen about half in length. Three specimens from a pair of humerus and 4 specimens from a pair of femur were immersed into 1N HCl from 10 minutes to maximum 70 minutes with 10 minutes’ interval for different level of demineralization. All 7 specimens were checked with BMD using by DXA (GE-Lunar Prodigy) and analysed respectively. The amount of calcium and phosphorus both from demineralized and normal area were measured and expressed in percentage of demineralizatioin. Results: As demineralization was going on with time of immersion into HCl, there was statistically significant correlation between the change of BMD and real change of calcium amount (?50.65) in humerus, and also in femur (?50.63). There was statistically significant correlation only in femur between BMD and phosphorus (?50.77), and not in humerus (?50.42). In summary, there was a high linear regression between BMD and real bone mineral with minimum of 89% and maximum of 97% as coefficient of determination (R-Square; R2) (p! 0.05). Through correlation analysis, correlation coefficient (?) between BMD measured by DXA and mineral of human long bone showed a high correlation as maximum of 0.84 (p!0.05). Conclusions: Our study showed a statistically significant high relationship between BMD measured by DXA and mineral of human long bone. Therefore, the measurement of DXA is considered to reflect the real change of mineral in human long bone as well as that of BMD. The limitation of this study was that it was based on artificially induced demineralization only one pair of cadaveric humerus and femur so that vertebrae were excluded.