A method and system for establishing an interactive media session based on IP Multimedia Subsystem, including: a terminal adapted to initiate an interactive media session request and receive a media session response; a serving-CSCF adapted to trigger the request to the application server according to a triggering rule and route the message; a proxy-CSCF adapted to forward the request and the response between the terminal and the serving-CSCF; an application server adapted to process the service request; a media control entity adapted to control resource allocation of the media carrier entity; and a media carrier entity adapted to allocate address ports for RTSP connection and RTP connection with the terminal. The present invention can be applied to a NGN network where the carrier and control are separated to improve the media delivery efficiency.