Access　control　has　been　widely　used　to　protect　sensitive　information　and　critical　resources　and　to　counter　malicious　attacks　in　computer　systems　and　networks.　There　are　three　main　traditional　access　control　models　in　the　past-discretionary　access　control　(DAC),　mandatory　access　control　(MAC)　and　role-based　access　control　(RBAC).　But　these　models　can　only　assign　rights　according　to　the　existing　access　control　policy,　which　is　stoical　and　reactive.　To　make　access　control　more　proactive,　some　experts　propose　some　access　control　models　based　on　game　theory.　But　complex　calculations　when　subjects　make　access　requests　will　affect　the　access　performance　and　reduce　the　access　efficiency.　To　solve　these　problems,　we　propose　an　access　control　model　based　on　stochastic　game　in　this　paper.　At　first,　we　introduce　a　strategy　selection　problem　of　object　mathematically　in　this　model,　and　then　describe　the　model　in　detail　on　how　to　solve　this　problem.　We　also　prove　the　existence　of　Nash　equilibrium　in　the　model.　At　last,　we　make　an　experiment　to　show　the　advantages　of　this　model.　We　prove　that　our　model　is　more　proactive　to　the　threats　and　attacks　in　network　than　traditional　access　control　models.