Abstract Animal models are widely used to explore the pathogenesis and management of osteoporosis. Mice are increasingly being used in animal models. We have evaluated the precision, accuracy, and ability to monitor changes in bone mineral measurements of mice with the Piximus and Hologic QDR 2000 devices. One hundred and twenty-two C57/BL6 mice were used in this study; 70 of them were put on a low calcium diet and followed prospectively for 14 wk. They were measured using both devices at baseline and at wk 14. Using the Piximus, we measured the whole body, the tibia, and two caudal vertebrae. Using the Hologic, we measured the tibia, which we divided into three equal parts. The remaining mice were used to evaluate the precision and accuracy of the measurement. The accuracy, which was determined only for the Hologic device, revealed a mean difference between the in vivo bone mineral content (BMC) and the ash weight of 0.1 mg. The precision, evaluated from the coefficient of variation (%) and the Smallest Detectable Difference (SDD, in absolute values) was good for both devices, confirming their ability to detect small differences in longitudinal studies: as little as 0.004 g for the BMC of the total tibia on both devices, and 0.003 g/cm 2 for whole body bone mineral density (BMD) on the Piximus. The BMC found using the two devices was comparable, whereas the BMD obtained on the Hologic device was nearly double that found using the Piximus. The comparison of the results by Bland and Altman's method showed that the difference between the results was not dependent on the magnitude of the measurement. We concluded that bone density and bone-density changes in mice can be measured precisely in vivo using the Hologic and Piximus devices; the latter being able to measure the whole body BMD with good precision.