The main goal of this paper is to propose a new methodological tool for economists, namely, agent-based computational models applied to Economics. There is a great opportunity after the severe criticisms by renowned economists and thinkers to the conventional economic theory due to its incapability to predict the last international financial crisis. Here, we describe the origins and main characteristics of this approach, and present four cases with relevant applications for economic theory: the theory of the firm, industrial clusters, asset valuation and financial bubbles, and the welfare theorems. This approach has the advantage of allowing us to model in a more realistic vein without loosing the necessary rigorousness to generalize results. Moreover, we can better integrate the microeconomic processes with the aggregate results out of the interaction of different agent types.
