With　the　increase　of　cars　in　China,　urban　traffic　congestion　is　becoming　a　serious　problem,　especially　during　the　morning　and　evening　rush　hours.　In　order　to　ease　congestion,　commuting　by　bus　has　been　proposed　as　a　feasible　alternative.　In　order　to　better　facilitate　travel　by　bus,　we　analyzed　the　commuter　travel　behavior　of　Tiantongyuan　community　in　Beijing.　Based　on　revealed　preference　(RP)　and　stated　preference　(SP),　travel　choice　behavior　can　be　obtained　with　the　actual　commuter　travel　conditions　and　the　commuter　travel　behavior　in　different　combinations　of　the　two　factors-travel　time　and　travel　cost.　A　binary　logit　model　(BL　model)　has　been　developed　based　on　SP　data,　and　with　this　future　traffic　patterns　or　policy　measures　can　be　predicted.　©　2012　American　Society　of　Civil　Engineering.