Amyloid-beta plaque deposition in the brain is one of the hallmark features of Alzheimer's disease (AD). Current estimation methods of amyloid burden depend on post-mortem histological analysis or resource-intensive contrast-enhanced imaging techniques. This dissertation focuses on developing a label-free, fast and non-invasive method based on spectral small-angle x-ray scattering (sSAXS) to estimate brain amyloid burden in vivo in small animals. Small-angle x-ray scattering (SAXS) is a well-established technique for identifying molecular structures based on their scattering features. However, SAXS is limited to the study of thin biological samples in the mm scale relying on low-energy monochromatic x rays. I built a prototype sSAXS system to tackle the sample thickness limit of traditional SAXS and explore in vivo applications with higher x-ray energies. This was achieved by integrating a polychromatic x-ray source with a 2D spectroscopic detector for simultaneously and efficiently collecting SAXS data in angle- and energy dispersive modes. A method based on sSAXS was introduced, and its capability was first demonstrated by identifying embedded targets in up to 5-cm-thick phantoms with an x-ray energy range between 30 and 45 keV. Wild-type and 5XFAD mice (AD animal model) were used to demonstrate the ability of the method to estimate amyloid burden in specific areas of the brain without using contrast agents. The mouse head was irradiated at selected locations using a two-pinhole collimated beam of polychromatic x rays for 300 s. The findings correlated well with the histological (gold standard) results. This work presents a promising new method based on sSAXS to estimate amyloid burden in the brain of small animals and possibly in humans.