The post-ovariectomy osteoporotic rat model is widely used to mimic post menopausal human osteoporosis and to test the efficacy of the therapies used in its treatment An experimental study was performed in order to improve the reliability of bone mineral density (BMD) measurements by correcting also the relocation error. In addition, a BMD map of the whole femur was carried out to detect the areas of uniform density, where the reliability of measuring was improved. Right femurs were taken from 5 Sprague Dawley female rats, 10 months old, and tested every 2 mm from the supracondylar line to the peritrochanteric line (6 scans for each site). Repositioning error was tested for each site of measurement, and measurements done by 3 different operators in a double blind test were compared. At site 4 and 24 (where the differences are high) and site 10 (where the differences of BMD are low) the coefficient of variability (CV) was calculated repositioning the bone after each measurement scan for a total of 6 scans. The CV at sites 4 and 24 were significantly higher than at site 10 (p < 0.05). In our opinion, the bone segment between 8 and 14 mm from the supracondylar line is that which presents the highest homogeneity and where it is preferable to perform the measurements in order to obtain the maximum effect of the method (improvement of precision of about 30%) in comparison with the other femur sites tested.