Satellite observations of nighttime light (NTL) play a vital role in monitoring urban dynamics. As NTL is observed from varying angles on a day-to-day basis, daily NTL data exhibits periodic fluctuations known as the angular effect. This phenomenon is believed to be closely linked to the three-dimensional (3-D) distribution of light sources. Therefore, our objective was to develop a model that can estimate 3-D urban structure based on NTL data by investigating the relationship between the angular effect and urban morphology. Firstly, we classified urban geometries using 3-D building model data and street data in Tokyo. Then, we compared these categorized urban structures with the observed angular effects. Our findings indicate a correlation between angular effects and specific urban structures, particularly in residential areas and districts characterized by dense high-rise buildings. The findings of this study provide a basis for using the angular effect to estimate urban structure. Expanding similar analyses to other cities will enable us to build a comprehensive model for urban structure estimation using NTL data. Such a model could provide more in-depth insights into urban land use, beyond the capabilities of traditional land cover classifications, especially in data-scarce regions.