As　urban　transportation　network　is　a　complex　system　with　various　traffic　facilities,　it　is　difficult　to　macroscopically　assess　its　performance.　Two-fluid　model　provides　a　measure　to　analyze　the　network-wide　characteristics　of　traffic　flow,　but　validation　of　the　model　needs　to　collect　the　traffic　data　of　individual　vehicles.　This　paper　suggests　a　reliable　and　feasible　method　to　evaluate　macroscopic　characteristic　of　urban　street　network　through　combination　of　GPS　(Global　Positioning　System)　and　Map-matching　based　on　GIS　(Geography　Information　System),　which　resolves　the　problems　in　data　collection.　The　application　of　this　method　redounds　to　the　improvement　of　urban　traffic　management　and　traveler　information　services.　Copyright　ASCE　2007.