Distribution　patterns　of　particulate　matter　2.5　(PM2.5)　inside　urban　street　canyons　varies　according　to　the　street　canyon　enclosure　ratio　(SCER)　and　traffic　flow　states,　and　therefore　they　affect　people　in　different　positions　of　street　canyons　to　different　extents.　This　paper　studied　PM2.5　distribution　patterns　in　both　symmetric　and　asymmetric　street　canyons　by　using　the　computational　fluid　dynamics　technique.　Simulation　experiments　were　conducted　in　a　series　of　isolated　street　canyons　in　Beijing　with　five　SCERs　and　three　types　of　traffic　flow　states.　The　results　of　the　simulation　experiments　were　verified　with　on-site　data.　The　results　are　threefold:　(1)　PM2.5　emitted　by　the　vehicles　gathers　around　the　corners　of　leeward　buildings　of　street　canyons,　and　this　phenomenon　tends　to　be　aggravated　by　heavy　traffic　flow　states　in　most　scenarios;　(2)　of　five　SCERs,　it　was　the　most　difficult　for　PM2.5　to　spread　out　of　the　canyons　with　SCER　=　2:1:2,　whereas　it　was　the　easiest　for　PM2.5　to　spread　out　of　the　canyons　with　SCER　=　2:2:1;　and　(3)　pedestrians　and　residents　on　the　leeward　side　of　asymmetric　street　canyons　are　exposed　to　the　highest　PM2.5　concentration.　The　findings　of　this　research　will　assist　in　designing　street　canyons,　controlling　the　traffic　flow,　and　developing　measures　to　reduce　the　harm　of　PM2.5　to　the　public.　(C)　2021　American　Society　of　Civil　Engineers.