In bone status assessment, proper quality assurance/quality control is crucial since changes due to disease or therapeutic treatment are very small, in the order of 2-5%. Unlike for dual X-ray absorptiometry, quality control procedures have not been extensively developed and validated for quantitative ultrasound technology, limiting its use in longitudinal monitoring. While the challenge of developing an ideal anthropometric phantom is still open, some manufacturers use the concept of the internal digital phantom mimicking human characteristics to check the stability of their device. The objective of the study was to develop a sensitive model of quality control suitable for the correction of QUS patient data. In order to achieve this goal, we simulated a longitudinal device lifetime with both correct and malfunctioning behaviors. Then, we verified the efficiency of digital phantoms in detecting those changes and subsequently established the in vitro/in vivo relationship. This is the first time that an attempt to validate an internal digital phantom has made, and that this type of validation approach is used. The digital phantom (DP) was designed to mimic normal bone (BUAP2) and osteoporotic bone (BUAP1) properties. The DP was studied using the UBIS 5000 ultrasound device (DMS, France). Diverse malfunctions of the UBIS-5000 were simulated. Several series of measurements were performed on both BUAP1 and 2 and on 12 volunteers at each grade of malfunction. The effect of each simulated malfunction on in vivo and in vitro results was presented graphically by plotting the average BUA values against the percentage change from baseline. The change from baseline in BUA was modeled using linear regression, and the in vivo/in vitro ratio was obtained from the model. All experimentations influenced the measure of BUAP1 and 2 as well as the measure of our 12 volunteers. However, the degree of significance varied as a function of the severity of the malfunction, and the results also differed substantially in magnitude between in vivo and in vitro. Indeed, the DP was about 10 times more sensitive to variations of the transfer function than was the in vivo measurement, which is very reassuring. The sensitivity of the digital phantoms was reliable in the determination of simulated malfunctions of the UBIS-5000. The digital phantoms provided an accurate evaluation of the acoustic performance of the scanner, including the fidelity of transducers. In light of these results, the QC approach of the UBIS-5000 will be extremely simple to implement compared with other devices. Indeed, since the digital phantom was automatically measured during every patient measurement, the QC approach could be built on an individual level basis rather than on an average basis.