Applications of vibrational spectroscopy to assess bone disease and therapeutic interventions are continually advancing, with tissue mineral and protein composition frequently investigated. Here, we used two spectroscopic approaches for determining bone composition in a mouse model (oim) of the brittle bone disease osteogenesis imperfecta (OI) with and without antiresorptive agent treatment (alendronate, or ALN, and RANK-Fc). Near-infrared (NIR) spectral analysis using a fiber optic probe and attenuated total reflection Fourier transform infrared spectroscopy (ATR FTIR) mode were applied to investigate bone composition, including water, mineral, and protein content. Spectral parameters revealed differences among the control wildtype (WT) and OIM groups. NIR spectral analysis of protein and water showed that OIM mouse humerii had ∼50% lower protein and ∼50% higher overall water content compared to WT bone. Moreover, some OIM-treated groups showed a reduction in bone water compared to OIM controls, approximating values observed in WT bone. Differences in bone quality based on increased mineral content and reduced carbonate content were also found between some groups of treated OIM and WT bone, but crystallinity did not differ among all groups. The spectroscopically determined parameters were evaluated for correlations with gold-standard mechanical testing values to gain insight into how composition influenced bone strength. As expected, bone mechanical strength parameters were consistently up to threefold greater in WT mice compared to OIM groups, except for stiffness in the ALN-treated OIM groups. Furthermore, bone stiffness, maximum load, and post-yield displacement showed the strongest correlations with NIR-determined protein content (positive correlations) and bound-water content (negative correlations). These results demonstrate that in this study, NIR spectral parameters were more sensitive to bone composition differences than ATR parameters, highlighting the potential of this nondestructive approach for screening of bone diseases and therapeutic efficacy in pre-clinical models.