Rainwater harvesting (RWH) has emerged as a promising technique to improve water security amid the escalating effects of climate change. However, a comprehensive evaluation of various rainwater harvesting solutions is needed to promote sustainable practices in the building sector. This study aimed to analyse rainwater harvesting in social housing in twelve representative cities in Brazil. Computer simulations were performed for 60 scenarios, comprising five social housing reference models and using rainfall data from twelve representative cities of Brazil’s bioclimatic zones. Simulation parameters of the building's characteristics were obtained from the blueprints of the house models, and rainfall data were obtained from Brazilian government databases. Other parameters, such as the runoff coefficient and the non-potable water demand, were retrieved from the existing literature. The results were analysed to understand the influence of the building’s characteristics and the different cities' rainfall patterns. Single-family houses have greater potable water savings potential (around 20% to 22% of total water consumption), primarily due to their higher roof-area-to-residents ratio. The multi-family models had a higher average total volume of potable water saved (from 108 to 152 m³/year) due to their higher total volume of water consumed. The city analysis revealed that Canela has the highest water saving potential (up to 35.8%), while Vitória da Conquista has the lowest average potential (0.35%).