This paper proposes a multi-agent framework for travel management decision support system that targets improved decision making, parallels human problem solving processes, and supports the major phases of decision making. A multi-agent system is a loosely coupled network of agents that work together to solve problems that are beyond the individual capabilities and knowledge of each agent. In travel management decision support system, agents are organized by their roles according to different phases of problem solving. They interact with the user, the environment, and each other, to improve the decision making process. Visual Prolog is used as our design language. The main advantage of using Visual Prolog is the flexibility Prolog rules provide in extending the functionality of agents.