With reference to the current issue of energy efficiency of residential buildings, the aim of this research is to analyze the possible influence of the energy performance component on the property prices formation. The study sample consists of two hundred residential units recently sold and located in the city of Bari (Italy). The implemented methodology is represented by a data-driven technique that employs a genetic algorithm to identify the functional expressions. The elaborations carried out have allowed the identification of a statistically reliable and easily interpretable model, which denotes an appreciable contribution of the energy component on housing prices.