To comprehensively evaluate the connectivity of urban gas network under different ground motion intensities, the method of seismic connectivity fragility analysis was proposed based on Monte Carlo simulation which uses the ground motion prediction equation (GMPE) as input. Firstly, the input ground motion intensity measurements (P_ga and P_gv) of each site of the gas network were determined based on GMPE conditioned on a given target magnitude. The distribution of random residual was simulated by normal distribution sampling to represent the uncertainty of the ground motion. Then, the Monte Carlo simulation was used to determine the failure probability of each element in the gas network system, and the loss index of connectivity was determined by the number of blocked nodes in the gas network after the earthquake. Finally, the seismic connectivity fragility curves of the gas network were obtained by calculating the loss index exceedance probability at different magnitudes. The gas network of a city in North China was taken as an example to evaluate its seismic connectivity fragility performance based on the proposed method, and the influence of uncertainty in GMPE was compared and analyzed. Calculation results show that the mean value of magnitude corresponding to each failure state of the gas network was 0.5 smaller when the uncertainty of ground motion was taken into account. Larger connectivity failure exceedance probability could also be observed at different magnitudes. The seismic connectivity fragility analysis method proposed in this study could provide constructive strategy for urban cities to evaluate the probability seismic risk of the whole gas network, and comprehensively take into account the seismic regional differences and uncertainties of ground motion.