Desert is a crucial component of terrestrial ecosystems, and its optical reflection properties are essential for understanding land surface radiative balance. Combining the hyperspectral and angular measurements has been considered as an effective method to characterize optical reflection properties of desert surfaces. However, the ground and satellite-based instruments cannot obtain both hyperspectral and angular reflectance of desert surfaces at appropriate scales due to the limitations of their measurement configurations. We employed an unmanned aerial vehicle (UAV) equipped with a spectrometer to perform hyperspectral BRDF measurements of six desert areas in Northwest China to overcome the limitations. In the comparison with those obtained from laboratory and satellite observations at the same geometries, it is found that UAV-measured hyperspectral reflectance has a spectral shape similar to that of spectra measured in laboratory settings, and UAV-measured angular reflectance shows a good match of BRDF characterization as the satellite observations in the hemispherical space. Moreover, we compared the modeled angular reflectance, which are derived from BRDF models, with measured angular reflectance of deserts at selected wavelengths from MODIS. The BRDF models using UAV-measured hyperspectral BRDF have superior performance in spectral BRDF simulation. The uncertainty analysis based on a Bayesian inversion further confirmed the reliability of the UAV data in effectively characterizing optical properties of the desert surface. These results demonstrate the robustness of UAV-based spectral and angular measurements in quantifying the hyperspectral BRDF, which provide comprehensive in-situ optical properties of desert surfaces. These data can serve as valuable validation for future studies on radiative balance of desert surfaces.