Carbon　emissions　trading　is　a　market-based　tool　for　solving　environmental　issues.　This　study　used　a　difference-in-differences　(DID)　approach　to　estimate　China＇s　carbon　trading　pilots　to　reduce　PM2.5　concentrations.　The　results　of　this　quasi-natural　experiment　show　that　the　carbon　trading　policy　effectively　reduces　PM2.5　by　2.7　mu　g/m(3).　We　used　a　propensity　score　matching　(PSM-DID)　method　to　minimize　selection　bias　to　construct　a　treatment　and　a　control　group.　The　results　show　the　policy　effect　is　robust,　with　a　PM2.5　concentration　reduction　of　2.6　mu　g/m(3).　Furthermore,　we　employed　a　series　of　robustness　checks　to　support　our　findings,　which　notably　indicate　that　the　effect　of　carbon　trading　on　reducing　PM2.5　differs　across　regions　over　the　years.　The　western　region　of　China　tends　to　be　the　most　easily　affected　region,　and　the　early　years　of　carbon　trading　show　slightly　greater　reduction　effects.　Our　findings　provide　valuable　policy　implications　for　establishing　and　promoting　carbon　trading　in　China　and　other　countries.