The molecular composition and hierarchical structure of bone tissue are important determinants of bone strength and fragility. Vibrational spectroscopic techniques have frequently been used to investigate bone composition, including attenuated total reflection (ATR) and Fourier transform infrared (FTIR) imaging spectroscopy. However, with these approaches, the spatial resolution of analysis is limited to several micrometers, which cannot capture properties at the nanoscale level of mineralized collagen fibrils (~500 nm), the building blocks of bone. Recent advances in optical photothermal infrared (O-PTIR) spectroscopy allows investigation of bone composition with submicron spatial resolution. In this thesis, we hypothesize that higher resolution information related to bone composition will provide new insights into factors underlying bone function. In two Aims, we: 1. Investigated whether O-PTIR-derived spectral parameters correlated to standard ATR spectral data from homogenized serially demineralized bone samples. 2. Examined whether O-PTIR-derived spectral parameters, including nanoscale heterogeneity of tissue, contribute to prediction of bone stiffness. Bovine tibial cortical bone was homogenized into powder and serially demineralized using EDTA. ATR and O-PTIR spectra were collected at six timepoints (n=6) over 5 days. Femoral neck bones from human donor cadaver tissues were sliced into 1mm sections. Line scans of 10 different trabecula were collected for each sample, with a total of 15 samples. Spectral images were collected from the intact femoral neck samples, which also had sex, age, and BMI information, as well as experimentally-determined stiffness parameters, associated with them. There was a correlation of R = 0.96 between the 2nd derivative phosphate (PO₄) /amide II peak intensity ratio for ATR to O-PTIR. Principal component analysis (PCA) yielded insight into the variance between O-PTIR and ATR spectra being mainly attributable to the phosphate absorbance peak (PO₄). We also found that O-PTIR nanoscale assessment of bone parameters was more sensitive to the acid phosphate peak (HPO₄) associated with newly formed bone. Partial least squares (PLS) regression analysis showed that combining multiple O-PTIR parameters (HPO₄ content and heterogeneity, collagen integrity and CO₃ content) can significantly predict whole-bone proximal femoral stiffness with an R² of 0.74. Overall, this study highlights a new application of O-PTIR spectroscopy to assess bone composition at a submicron scale, which may provide new insights into molecular-level factors underlying bone mechanical competence.