In order to improve the overall network operation efficiency of urban rail transit and its service，an in- dex which measures the integrated importance of transfer nodes is put forward. Based on this，a coordinated opti- mization model and algorithm of network operation schedule are established with the consideration of passenger demand. This model aims at reducing the passenger transfer waiting time and facilitating the transfer of passen- gers in the transfer station. Finally，the proposed method and algorithm are applied in a real rail transit to vali- date the effectiveness. The results show that after the coordination step by step，the average waiting time of the whole network has been reduced by 12.42%，saving 0.39 minutes for passengers. It proves that the model and al- gorithm can effectively reduce passenger transfer waiting time，optimize the coordination of network operation schedule，and improve the transfer efficiency of the whole network.