Shaded objects in remote sensing data are usually neglected because of their low reflectance. The common treatment for detecting shadows consists of simple identification and straightforward removal; however, this approach is often criticized owing to the complexity and difficulty of its application and the incompleteness of the treated data. Recent efforts with hyperspectral field reflectance measurements have provided alternatives to scrutinize the issues of shadows. To demonstrate our proposed framework, we first gathered basic hyperspectral data with a spectrometer at Peking University, Beijing, China. Subsequently, we classified the spectral reflectances into three types: sunlight reflectance, shadow reflectance, and simulated reflectance. The spectral characteristics of shadows can provide pertinent classification information despite their low reflectance. Under full shadow conditions, the simulated reflectance was close to the sunlight reflectance because the same reduction ratio was used for the spectralon panel during the measurement process. In addition, the ratio of the shadow reflectance to the sunlight reflectance seemed to be related to the characteristics of the obstruction. The findings in this paper can contribute to a comprehensive understanding of the influence of shadows in remote sensing research, demonstrating the potential of fully utilizing typically neglected shadow information.