In urban environments, decentralized energy systems from renewable photovoltaic resources, clean and available, are gradually replacing conventional energy systems as an attractive source for electricity generation. Especially with the availability of unexploited rooftop areas and the ease of installation, along with technological development and permanent cost reductions of photovoltaic panels. However, the optimal use of these systems requires accurate estimates of supply (rooftop solar photovoltaic potential) and the design of an intelligent distributed-system integrated with power grids. Geographic information systems (GISs)-based estimation is justified as a promising approach for estimating rooftop solar photovoltaic potential, in particular, the possibility of combining GISs with LiDAR (Lighting-Detection-And-Ranging) to build robust approaches leading to accurate estimates of the rooftop solar photovoltaic potential. Accordingly, this study aims to present a comprehensive review of GISs-based rooftop solar photovoltaic potential estimation approaches that have been applied at different scales, including countries. The study classified GISs-based approaches into sampling, geostatistics, modeling, and machine learning. The applications, advantages, and disadvantages of each approach were reviewed and discussed. The results revealed that GISs-based rooftop solar photovoltaic potential estimation approaches, can be applied to the large-scale spatial-temporal assessment of future energy systems with decentralized electrical energy grids. Assessment results can be employed to propose effective-policies for rooftop photovoltaic integration in built environments. However, the development of a new methodology that integrates GISs with machine learning to provide an accurate and less computationally demanding alternative to LiDAR-based approaches, will contribute significantly to large-scale estimates of the solar photovoltaic potential of building rooftops.