Because　of　complex　facility　layout　and　large　number　of　passengers,　it　is　quite　difficult　to　monitor　passengers　in　subway　transfer　stations.　Artificial　identification　is　still　the　common　method　to　find　out　pedestrian　congestion　and　other　abnormal　situation.　Thus,　in　order　to　improve　the　efficiency　of　passenger　monitoring,　it　is　necessary　to　identify　key　nodes　to　support　the　decision　making　of　monitoring　equipments　configuration　and　operation.　An　operational　strategy　for　key　monitor　nodes　identification　using　grey　relational　analysis　(GRA)　is　presented.　Passenger　facilities　were　divided　into　four　types　to　draw　potential　monitoring　nodes　diagram.　And　then,　based　on　the　pedestrian　simulation　tool,　evaluation　indicators　system　of　monitor　node　importance　was　established.　GRA　algorithm　with　variable　weight　was　used　to　calculate　importance　of　different　potential　monitoring　nodes.　At　last,　the　application　of　the　identification　method　was　illustrated　with　a　case　study　of　a　designing　comprehensive　transportation　hub　in　Beijing.