During the last decade a new modeling approach, known as agent-based modeling, was developed and made its way into several areas of research. It greatly benefited from rapid developments in computing technology. As for economics, agent-based modeling is based on the observation that social systems can be understood and modeled explicitly as systems of interacting agents. Because of their flexibility with respect to the representation of agents, the interactions between agents as well as the resulting dynamic and spatial relationships, these models offer a great variety of perspectives for future research. In this contribution, the underlying modeling concept is presented, and exemplary applications to the field of sector and policy analysis are discussed. Future perspectives of agent-based modeling are pointed out, but also challenges to the agent-based approach are presented.