The invention discloses a network virus transmission behavior modeling method. The method comprises a network initialization process, a virus transmission process, a node detection process, an immunization and immunization failure process, a model evaluation process and an immunization strategy proposing process. When a model runs, an infected state (I) node transmits a virus to a neighbor node at certain probability; an infectious state (S) node is infected into the infected state (I) node by the virus at the certain probability; meanwhile, the infected state (I) node is isolated at the certain probability, and becomes an isolated state (Q) node which does not have the virus transmission capacity and cannot be infected; and an immune state (R) node which cannot be infected by the virus appears in the model through treatment on the isolated state (Q) node and immunization on all nodes. The factors affecting virus transmission in the SIOR model include total node number n, average node degree <k>, initially infected node number n<0>, virus transmission rate alpha, isolating treatment rate beta and isolating recovery rate lambda; and the cured node has immune efficacy probability mu, vaccine injection rate gamma and vaccine failure rate eta. An achievable immunization strategy is proposed on the basis of these factors.