A　coupled　MM5-ARPS-CMAQ　air　quality　modeling　system　was　applied　for　investigating　the　contributions　of　various　emission　sources　to　the　ambient　concentrations　of　PM10　in　the　Beijing　metropolitan　region　of　China.　A　three-level　nested　simulation　domain　was　established　for　facilitating　the　modeling　study,　and　a　GIS-based　atmospheric　pollutant　emission　database　was　constructed　to　provide　important　emission　information　for　the　Models-3/CMAQ　model.　After　verification,　the　coupled　modeling　system　was　implemented　to　predict　the　hourly　contribution　ratios　of　various　PM10　emission　sources　to　Beijing＇s　air　quality　during　four　representative　months　of　January,　April,　August　and　October　in　the　base　year　of　2002.　It　was　found　from　the　modeling　results　that　three　major　sources　(including　fugitive　industrial　emissions,　construction　sites,　and　road　dusts)　presented　a　relatively　high　contribution　to　the　PM10　pollution　in　Beijing,　with　the　contribution　ratios　of　28.67%,　42.88%,　42.46%　and　35.67%　for　spring,　summer,　autumn　and　winter,　respectively.　In　addition,　the　impacts　of　emissions　from　two　major　industrial　zones　within　Beijing　on　its　air　quality　were　quantified.　Moreover,　the　coupled　modeling　system　was　applied　to　explore　possible　future　PM10　concentration　distributions　in　2008　through　two　proposed　emission　sources　reduction　strategies.　It　was　found　that　Beijing　needs　to　take　continual　emission　control　efforts　to　comprehensively　improve　its　air　quality　to　the　Class　II　level　by　2008.　The　modeling　results　could　provide　sound　decision　making　basis　for　developing　effective　air　quality　management　strategies　in　the　planning　region.　(C)　2006　Published　by　Elsevier　Ltd.