 This paper presents a new methodology for determining the optimal storage capacity of a rainwater harvesting system based on probability theory. The proposed model takes into consideration the variability of rainfall, the size of the demand, and the number of consecutive dry periods. It provides a way to calculate the required storage capacity with a given probability of not exceeding a certain amount of water. The model was tested using data from Milan, Italy, and the results were found to be accurate. This article was authored by Maria Gloria Di Ciano, Mariana Marioni, Anita Ramondi, and others.