With　the　development　of　multimodal　transportation　systems,　travelers　face　various　choices　to　decide　how　to　travel　from　one　place　to　another.　Given　uncertainty,　former　experience,　and　individual　characteristics　such　as　income　level　and　risk　preference,　a　multiround　decision-making　model　based　on　cumulative　prospect　theory　(CPT)　is　proposed.　This　model　was　tested　through　a　series　of　behavioral　experiments　conducted　on　the　oTree　platform,　which　involved　participants　making　travel　mode　choices　across　multiple　iterations.　The　experiment　was　designed　with　three　income　levels:　low,　medium,　and　high;　three　risk　preferences:　risk　aversion,　risk　neutral,　and　risk　lover;　and　two　scenarios:　under　urgent　event　and　nonurgent　event.　The　experiments　were　successful　in　highlighting　the　influence　of　previous　experience　on　mode　choice,　with　subjects　displaying　varying　degrees　of　sensitivity　to　tardiness　and　33%　of　subjects　displaying　insensitivity.　The　data　obtained　from　the　experiment　were　used　for　model　fitting.　Then　the　effects　of　travel　time,　fare,　and　travel　time　reliability　on　different　income　groups　were　estimated　under　various　situations.　The　reference　points　of　CPT　were　set　based　on　heterogeneous　groups　and　adjusted　each　round　to　achieve　a　subjective　and　dynamic　effect.　The　results　show　that　in　urgent　scenarios,　medium-income　groups　have　the　highest　value　of　time　(VOT)　of　5.74　RMB/min.　Moreover,　high　incomes　have　the　value　of　reliability　(VOR)　of　-19.52　RMB/min,　indicating　they　are　willing　to　pay　an　additional　cost　of　19.52　RMB　per　minute　for　choosing　a　reliable　mode.　These　results　facilitate　discussions　on　how　to　guide　travelers　toward　green　travel　modes　and　what　strategies　enterprises　can　adopt　to　attract　travelers.　This　model　has　the　potential　to　enhance　urban　traffic　management　and　policy　development　tailored　to　diverse　traveler　groups.