In　order　to　explore　the　establishment　method　of　high-resolution　vehicle　emission　inventory　and　analyze　the　temporal　and　spatial　variation　law　of　vehicle　greenhouse　gases　(GHG)　emissions.　This　paper　proposes　a　bottom-up　method　based　on　road　network　information　and　the　real-time　average　interval　speed　of　road　segments.　A　traffic　speed-flow　model　is　proposed　to　predict　the　hourly　traffic　flow　and　the　localized　Motor　Vehicle　Emission　Simulator　(MOVES)　is　used　to　simulate　the　CO2,　N2O　and　CH4　emission　factors.　A　high　temporal　(1　h　x　1　h)　and　spatial　(1　km　x　1　km)　resolution　GHG　emission　inventory　of　motor　vehicles　in　Beijing　in　2018　is　developed　by　this　means.　The　actual　emissions　of　CO2,　N2O　and　CH4　are　19,864,590,　82　and　511　t,　respectively.　And　the　total　GHG　emission　is　19,901,933　tCO(2e)　combined　with　global　warming　potential　(GWP).　The　daily　GHG　emissions　on　weekday　and　weekend　are　55206.30　and　52817.64　tCO(2e),　respectively.　There　are　three　obvious　peak　emission　periods　on　the　weekday,　namely,　the　morning　peak　(8:00-9:00),　the　noon　peak　(11:00-12:00)　and　the　evening　peak　(18:00-19:00),　which　contribute　11.76%,　11.84%　and　12.92%　of　the　daily　emission,　respectively.　The　1　h　x　1　h　emission　grid　shows　the　spatial　distribution　characteristics　of　emissions.　The　areas　within　the　Fifth　Ring　Road　(973　km(2))　are　only　5.93%　of　the　total　area　of　the　city　(16,410　km(2))　but　contribute　41.53%　of　the　total　vehicle　GHG　emissions.　This　study　provides　detailed　data　support　for　implementing　vehicle　GHG　emission　mitigation　measures.　(C)　2020　Published　by　Elsevier　Ltd.