This research effort focuses on the development of a unified, systematic, and formal approach to modeling and control of multi-agent systems, drawing inspiration from biological systems exhibiting swarm behavior. Previously developed works in the application of graph theory were further advanced and applied to a variety of different task-oriented multi-agent systems. New distributed on-board controllers were developed and implemented at the individual level, yielding collective results in exploration, perimeter detection and surveillance, and formation control that are indistinguishable from similar centrally controlled systems. This distributed control yields robustness and decreased complexity relative to centralized controllers. Simulations and experiments were conducted to validate the analytically obtained results. The major outcomes of this research have been published in a large number of peer-reviewed conference proceedings and scientific journal articles. The financial support provided by the Army Research Office to carry out this research has been duly acknowledged in these articles.