Over the last quarter century it has become increasingly vital for large-scale businesses to exploit revenue management strategies. This is a consequence of thinning profit margins that have diminished from the effects of several pervasive trends. Among these trends are the globalization of markets, the regulation of particular industries, and, especially in the airline industry, the introduction of heightened security standards. To keep pace with this volatile landscape, revenue management and order promising policies rapidly have become more complex. Correspondingly, testing such policies has become increasingly difficult. This thesis presents an extensible test-bed designed to facilitate testing complex revenue management and order promising policies. The testbed was modeled and built using an agent-oriented architecture in order to reflect real-life business environments accurately. We describe the design and construction of the test-bed in this thesis. We then present case studies to highlight its capabilities.