With　the　boom　of　tourist　travel　recently,　the　frequency　of　leisure　activities　is　increasing,　which　results　in　rising　of　tourism　traffic.　Tourist　travel　brings　great　pressure　to　urban　roadway　network,　highways　as　well　as　attractions.　Many　traffic　challenges　emerged,　such　as　traffic　jams,　parking　difficulties,　etc.;　these　problems　greatly　reduced　tourists　travel　satisfaction.　In　this　study,　investigations　were　conducted　to　make　sure　the　travel　characteristics　of　three　kinds　of　scenic　spots　including　the　landmark,　natural　beauty　and　cultural　heritage　sites.　Factors　which　have　an　effect　on　choice　of　traffic　modes　and　destinations　were　analyzed　including　personal　property,　travel　features,　traffic　states,　traffic　mode　preferences　and　tourist　traffic　management　measures.　Binary　Logit　(BL)　regression　was　used　to　set　up　model　of　traffic　mode　choice.　The　key　factors　which　affect　traffic　mode　choice　were　acquired　from　the　BL　model,　including:　travel　scale,　visit　times,　per　capita　daily　consumption,　age,　private　cars　possession　and　attractions　types.　©　2011　ASCE.