To　meet　the　demand　for　high-resolution　air　quality　study　over　the　complex　terrain　in　Beijing　region,　the　Advanced　Regional　Prediction　System　(ARPS)　was　introduced　to　replace　the　default　meteorological　driver　in　Models-3　Community　Multiscale　Air　Quality　(CMAQ)　modeling　system,　MM5,　to　provide　a　complete　set　of　meteorological　data　needed　for　air　quality　simulations.　The　ARPS　model　was　chosen　because　its　non　hydrostatic　dynamics,　its　generalized　terrain　following　height　vertical　coordinate　and　Arakawa-C　horizontal　grids　which　are　identical　to　the　ones　used　in　CMAQ,　and　its　outstanding　capabilities　of　simulating　small　scale　circulations　as　well　as　weather　systems　at　other　scales.　It　is　well　suited　for　the　complexities　of　the　Beijing　region.　Based　on　the　Meteorology-Chemistry　Interface　Processor　(MCIP)　in　the　CMAQ　system,　this　paper　describes　the　process　and　software　components　used　to　link　the　ARPS　model　and　the　CMAQ　system.　In　addition,　the　coupled　modeling　system　was　verified　through　plotting　scatter　diagrams　and　analyzing　Pearson　correlation　coefficient　between　simulated　PM10　concentration　and　observation　data　provided　by　three　selected　monitoring　stations　for　August　in　2002.　The　results　show　that　a　satisfactory　agreement　has　been　reached　between　modeled　and　observed　data.　This　not　only　indicates　acceptable　model　prediction　accuracy　and　model　performance　has　been　reached,　more　importantly,　it　stands　for　an　option　of　modeling　system　which　can　help　us　to　investigate　the　air　quality　over　the　area　of　complex　topography　has　been　established.