Abstract: An in-depth analysis is made of the interdependence concerning the information security. The priority is given to the study of the relationship between the cost and the expected benefits from the security loss among different agents. And the agent's choice of strategy can be the following two: to invest into security cost; or not invest into security cost. By means of the pay-off matrix, a game theoretical model is established for one agent or more agents. This model uses Nash equilibrium analysis to provide decision support for the agent's cost investment. When the problem of how to apply the information security cost investment to improve our own security is solved, effort is made to conduct the study of the cost relationship between the various network agents and it is suggested that the “equilibrium” in the game theory be used to quantify the relevant factors of the various agents who choose to invest into information security cost, thus to transform the model into a relative multiplayer game model so that a theoretical approach can be worked out for solving the problem of information security cost investment thus to achieve the effective risk minimization at a limited cost.