There　are　two　drawbacks　in　the　existing　methods　on　the　stop　selection　of　express　bus　services.　One　is　that　the　pivotal　role　of　bus　stops　is　ignored,　and　the　other　is　that　the　methods　usually　only　consider　bus　passenger　flow　data　but　does　not　consider　other　data　sources　such　as　taxi.　By　introducing　the　complex　network　theory,　this　paper　integrates　the　bus　and　taxi　passenger　flow　data,　and　proposes　a　recommendation　method　of　limited-stop　bus　stops　based　on　multi-criteria　collaboration　which　comprehensively　considers　the　passenger　volume　at　bus　stops,　potential　passenger　flow　and　the　pivotal　role　of　bus　stops.　Firstly,　based　on　the　data　of　bus　passenger　flow,　bus　lines　and　taxi　passenger　flow,　we　calculate　the　bus　passenger　volume　and　potential　passenger　volume.　Secondly,　the　importance　degree　of　bus　stops　is　calculated　by　constructing　a　directed-weighted　complex　network　model　of　the　urban　bus　system.　Finally,　the　important　degree　of　bus　stops,　passenger　volume　and　potential　passenger　volume　are　weighted　and　summed　to　gain　the　comprehensive　score　of　bus　stops　by　analytic　hierarchy　process,　and　the　recommendation　method　of　the　bus　stop　combination　is　constructed.　Taking　Xiamen　as　an　example,　the　method　is　applied　to　conduct　the　bus　stops　recommendation.　The　results　show　that,　with　the　introduction　of　the　stop　importance　and　potential　passenger　volume,　the　recommended　stops　are　more　suitable　for　the　actual　resident　travel　demand　compared　with　the　traditional　methods.　Copyright　©　2021　by　Science　Press.