The urban rail transit network is an important part of an urban public transportation system. First, we generated the network models of the urban rail networks of Beijing and Shenzhen. We used the subway smart card data to estimate the passenger travel demands in the two urban rail transit networks. Next, we analyzed the topological structures of the two urban rail transit networks based on complex network theory and proposed the indices to evaluate the vulnerability of an urban rail transit network. Finally, we generated a two-layer public transportation network to obtain a deeper understanding of the vulnerability of urban rail transit networks. Our empirical results show that the distribution of vulnerable segments is similar in Beijing and Shenzhen rail transit networks. Averagely, Shenzhen urban rail transit network is more vulnerable. The vulnerability of an urban rail network is highly related with its network complexity. Urban bus transit network can reduce the vulnerability of urban rail transit system.