This　paper,　which　aims　at　the　primary　gaseous　air　pollutants　(i.e.,　SO2,　NOx,　VOCS　and　CO),　is　the　third　paper　in　the　series　papers　published　in　Atmospheric　Environment　to　develop　new　emission　estimation　models　by　the　regression　method.　A　group　of　regression　models　for　various　industrial　and　non-industrial　sectors　were　proposed　based　on　an　emission　investigation　case　study　of　Handan　region　in　northern　China.　The　main　data　requirements　of　the　regression　models　for　industrial　sectors　were　coal　consumption,　oil　consumption,　gaseous　fuel　consumption　and　annual　industrial　output.　The　data　requirements　for　non-industrial　sector　emission　estimations　were　the　population,　the　number　of　resident　population　households,　the　vehicle　population,　the　area　of　construction　sites,　the　forestland　area,　and　the　orchard　area.　The　models　were　then　applied　to　Tangshan　region　in　northern　China.　The　results　showed　that　the　developed　regression　models　had　relatively　satisfactory　performance.　The　modeling　errors　at　the　regional　level　for　SO2,　NOx,　VOCS　and　CO　were　-16.5%,　-10.6%,　-11.8%　and　-22.6%,　respectively.　The　corresponding　modeling　errors　at　the　county　level　were　39.9%,　33.9%,　46.3%　and　46.9%,　respectively.　The　models　were　also　applied　to　other　regions　in　northern　China.　The　results　revealed　that　the　new　models　could　develop　emission　inventories　with　generally　lower　error　than　found　in　previous　emission　inventory　studies.　The　developed　models　had　the　advantages　of　only　using　publicly　available　statistical　information　for　developing　high-accuracy　and　high-resolution　emission　inventory,　without　requiring　detailed　data　investigation　which　is　necessary　by　conventional　＂bottom-up＂　emission　inventory　development　approach.　(C)　2014　Elsevier　Ltd.　All　rights　reserved.