In content-based unstructured P2P search system, the main issues that affect query efficiency and searching cost are the complexity of computing document similarity brought by high dimensions and the great deal of redundant messages coming with flooding. Content-based cluster P2P search model depending on set distance is proposed in this paper to reduce the query time and redundant messages. This model defines document similarity by set distance to restrain the complexity of computing the document similarity in linear time. Also, clustering peers based on content depending on set distance reduces the query time and decreases the redundant messages. Simulations show that this model not only has higher recall, but also reduces the search cost and query time to the rate of 40% and 30% of Gnutella.