A　calculated　methodology　of　regional　contributions　based　on　the　MM5-CAMx　model　was　developed　to　analyze　the　impact　of　different　emission　regions　on　urban　air　quality.　Meteorological　fields　of　January,　April,　July　and　October　in　2006　were　obtained　by　running　the　MM5　modeling　system.　The　SO2　source　apportionment　technology　was　used　to　investigate　the　influence　of　the　regions　on　air　quality　in　Tangshan.　A　K-means　clustering　algorithm　was　used　to　group　these　contribution　rates　into　five　clusters　according　to　daily　average　concentration　contribution　of　120　days.　Results　indicate　that　the　receptor-self　as　emission　region　has　the　important　contribution　to　receptor　sites,　the　pollution　could　be　aggravated　by　local　source　emissions.　Combined　with　the　Tangshan　emission　inventory,　the　analysis　shows　that　several　industries　are　located　in　inappropriate　sits.　Additionally　the　SO2　and　PSO4　source　apportionment　study　reveals　a　few　differences　from　the　contributions　of　emission　regions.