Cloud Computing technology enables the sharing and collaborating of wide variety of resources. To fully utilize these resources, effective discovery techniques are necessities. Proposing and designing a resource discovery scheme based on Economic Agent. Base on the economic model and the technique in agent of grouping nodes sharing similar files to improve efficiency, this thesis suggests a resource discovery scheme based on economic agent, which is called EAGRD. Theoretical models on resource discovery are provided, under which EAGRD is compared with existing schemes theoretically. By controlling propagation of message into related communities, EAGRD improves time and network efficiency at the cost of topological maintenance overhead. Results from simulation demonstrate that this architecture is very effective in Cloud Computing resource discovery.