This paper describes the use of artificial intelligence, as soft computing methods based on artificial neural network (ANN) for estimating repair and maintenance works of buildings. The use of artificial neural network in estimating repair and maintenance of buildings has shown its ability, usefulness and preciseness. Research into accurate models for estimating the repair and maintenance of buildings has led to the development of models based on artificial neural network. The results show that the accuracies obtained range within 4% and 6%, which lend to the view that repair and maintenance costs for masonry buildings are less than those for reinforced concrete buildings.