This　paper　presents　a　disaggregate　mode　choice　model　which　attempts　to　reveal　the　inherent　laws　interacting　between　transportation　and　land　use　as　well　as　find　solutions　for　the　urban　traffic　congestion　problem.　The　Genetic　Algorithm　method　and　Monte　Carlo　simulation　are　adopted　for　model　parameter　estimation.　2006　Beijing　travel　survey　data　is　used　for　model　validation.　The　empirical　results　show　that　all　parameters　have　the　expected　signs　with　the　goodness　of　fit　equal　to　0.759.　Each　mode　except　for　the　passenger　mode　achieved　80.9%,　95.1%,　88.7%,　88.3%　and　70.3%　of　successful　predictions　for　the　auto,　bike,　transit,　walk　and　taxi　respectively.　The　Genetic　Algorithm　based　disaggregate　mode　choice　model　is　able　to　effectively　reflect　the　intrinsic　links　between　urban　transportation　and　land　use.　©　2013　WIT　Press.