Taxis　account　for　a　large　percent　of　urban　public　passenger　transport.　This　research　proposes　a　monitor　model　of　analyzing　unoccupied　taxis　at　a　defined　area.　By　scientific　analysis　of　taxi　GPS　data　precision　and　quality,　16,255　valid　taxis　in　register　list　were　filtered,　and　two　coefficients　were　introduced　to　correct　the　taxi　monitoring　model.　Based　on　the　location　reports　of　valid　taxis,　the　number　of　unoccupied　taxi　which　was　able　to　dispatch　to　meet　passengers’　demands　can　be　predicted.　Take　the　terminal　3　of　Beijing　Capital　International　Airport　as　a　case　study,　the　field　investigation　results　show　that　the　precision　of　model　reached　87.25%.　It　provides　reliable　systematic　support　and　alert　for　a　low　taxi　number　when　there　comes　up　large　taxi　travel　demand　during　the　peak　period　or　under　severe　weathers.