The　supplier　selection　plays　an　important　role　in　supplier　chain　management.　How　to　evaluate　the　performance　of　suppliers　is　still　an　open　issue.　Multi-criteria　decision-making　(MCDM),　due　to　its　ability　of　solving　multi-source　information　problem,　has　become　a　quite　effective　tool.　Currently,　the　analytic　network　process　(ANP)　and　Entropy　weight　are　employed　to　solved　MCDM　problems.　However,　these　techniques　ignore　the　one-sidedness　of　the　single　weighting　method　and　cannot　deal　with　the　uncertainties　of　input　data.　In　this　paper,　a　new　evidential　ANP　methodology　based　on　game　theory　is　proposed　to　efficiently　address　supplier　management　under　uncertain　environment.　First,　ANP　and　entropy　weight　are　employed　to　obtain　the　subjective　and　objective　weights　of　criteria.　Second,　based　on　decision-making　trial　and　evaluation　laboratory　(DEMATEL)　and　game　theory,　the　comprehensive　weight　of　ANP　and　entropy　weight　can　be　determined.　Game　theory　is　employed　to　combine　the　merits　of　subjective　weight　and　objective　weight,　and　DEMATEL　is　adopted　to　adjust　the　weight　of　criteria　to　make　the　result　more　reasonable.　Finally,　evidence　theory　is　adopted　to　deal　with　the　uncertainties　of　input　data　and　get　the　supplier　selection　result.　A　case　study　is　given　to　demonstrate　the　proposed　modeling　process.　By　comparing　with　the　existing　methods,　we　demonstrate　that　the　proposed　model　has　many　advantages　and　it　shows　the　efficiency　and　rationality　in　supplier　selection　problem.