The invention discloses a method for identifying false evaluations and preventing malicious attack in a P2P network, which comprises the following steps: distributing file preservation nodes and trust management nodes for each file in a network; recording detailed information of each download process by the file preservation nodes; when the trust management nodes receive new evaluation information, starting an evaluation authenticity detecting process; verifying whether the node for reporting evaluation downloads the evaluated file or not by asking a file preservation mode download record, collecting download process details and asking challenge problem, and the like, thereby judging whether the evaluation is credible or not, meanwhile, estimating the node reporting the estimation; and adopting a corresponding verifying strategy for the reported evaluation according to a history estimation record of the node. Compared with the prior art, the method has wider application range, can more effectively detect the authenticity of estimation actions in the P2P network and greatly increase the resistance of a P2P network trust model for the attack.