To　explore　spatiotemporal　characteristics　of　vehicles　in　a　long　tunnel　road　under　the　vehicle-road　coopera-tive　environment,　the　human-machine　interface　of　tunnel　warning　system　based　on　the　connected　vehicle　technology　was　designed　and　a　driving　simulation　experiment　was　carried　out.　35　drivers＇　spatiotemporal　characteristics　of　vehicle　were　studied　by　taking　the　group　without　warning　system　as　the　control　group.　The　results　reveal　that　when　the　system　was　on,　the　slowdown　trigger　points　of　vehicle　was　earlier,　and　the　curvature　values　and　number　of　inflection　points　of　the　spatiotemporal　graph　decreased.　Moreover,　the　coefficient　of　variation　of　vehicle　speed　near　the　entrance　and　inside　tunnel　and　the　degree　of　overspeed　in　tunnel　decreased　obviously　when　the　system　was　on.　In　conclusion,　with　the　warning　system,　the　driving　pattern　was　changed　from　a　visual　stimulation-based　response　behavior　to　a　psychological　expectation-based　proactive　behavior,　which　was　easy　to　form　the　stable　spatiotemporal　characteristics.　Further　research　shows　that,　assuming　that　the　minimum　headway　of　the　following　distance　is　1.5s,　and　based　on　the　best　converging　patterns　from　35　drivers＇　spatiotemporal　trajectories,　the　overall　traffic　capacity　was　21.27%　higher　than　that　without　the　warning　system.　©　2020,　Editorial　Department,　Journal　of　South　China　University　of　Technology.　All　right　reserved.