Abstract The prevailing High-rise Residential Buildings (HRBs) alleviate land shortage in modern cities, but energy sustainability in these energy-intensive constructions remains a big problem. The large available surface area of HRB makes rooftop attached PV technique a competent solution to utilize the local renewable energy. In the future, this can be expanded through facade installations. However, to which degree could PV generation support the HRB electricity consumption remains unclear. With more variations coming from renewable energy sources to the grid, peak-load electricity reduction will become critical. In this paper, the peak load shaving potential of rooftop attached PVs in HRB is estimated. Multi-agent system framework is employed to identify the building and occupant characteristics of HRB electricity consumption. A series of comparisons between demand and net demand profiles of HRB in a typical summer week are conducted. The results show that PV could effectively reduce the peak load during daytime, however, the maximum PV generation in the midday may cause new valleys in the net demand profiles. The results suggest that energy storage should be used to optimally match demand and supply in order to smoothen the net demand profiles.