Energy consumption of buildings worldwide has steadily increased over the past couple of decades. Furthermore, energy performance of buildings is one of the factors that contribute to energy waste and its perennial adverse impact on the environment. This paper presents a data mining approach for assessing the heating and cooling requirements of residential buildings. The proposed approach combines Artificial Neural Networks (ANNs) and cluster analysis to assess and predict the heating and cooling energy efficiency of residential buildings. The ANN-based model uses eight input variables (i.e., relative compactness, surface area, wall area, roof area, overall height, orientation, glazing area, and glazing area distribution) to predict both the heating and cooling loads of residential buildings. Buildings are then clustered based on the output variables using the K-means clustering method. The proposed approach is used to assess and evaluate 768 diverse residential buildings based on simulated literature data. The research results showed that the proposed approach can effectively predict the heating and cooling requirements of residential buildings based on the input variables considered with a very high level of accuracy.