Abstract: In order to solve the limitations of slow network response and more manual intervention in the existing intrusion detection system,a new method of network intrusion detection is proposed based on agent technology and clustering analysis.The distributed intrusion detection system is designed with the agent technology,in which the system is self-controlled by the hierarchical control agent,and the data analysis of network is achieved through data mining technique based on fuzzy C-means algorithm(FCM).The weighted algorithm is used in the FCM algorithm for the improvement of the system detection capabilities.The results show that the system is able to reduce manual intervention,achieve faster response for the network environment,and improve the performance of the intrusion detection.