Abstract: Velocity pulses and long-period response spectrum are typical features of near-fault ground motions, resulting in a strong seismic response of the cable-stayed bridges than the far-field ground motions. To quantify the contribution of velocity pulses and response spectrum characteristics to the seismic response of the cable-stayed bridges, this paper employs the pulse-type ground motions synthesis method that can match the target response spectrum to obtain random ground motions. Based on the validity of the synthetic ground motions, takes a cable-stayed bridge as an example, the influence of the velocity pulses and response spectrum characteristics on the longitudinal stochastic seismic response of cable-stayed bridges near faults is investigated using the Monte Carlo simulation, respectively. The results indicate that the maximum sectional curvature of the bridge towers in the studied example is the most sensitive to the velocity pulse and the response spectra of near-fault ground motions and the sensitivity is similar. The mean values of the response are approximately 50% larger than those of the two groups of ground motions without velocity pulses and with the normative spectra as the target spectra. The internal forces at the bottom of the piers are affected by the response spectrum and the velocity pulses approximately with the mean values of the response increases by 25%, which is lower than that of the bridge tower. The response spectrum characteristics have a greater impact on the displacement of the main girder than that of the velocity pulses, with the difference of 17%, due to the long-period part of the response spectrum. The amplification of the dynamic response of cable-stayed bridges under pulse-type ground motions is caused by both the response spectrum characteristics and the time-domain velocity impulse. There is a possibility of bias in insecurity in the analysis of the seismic response of the bridge structure by selecting the measured records according to the matching principle of the normative spectra as the target spectra. The response spectrum characteristics should be considered as an important analytical factor in addition to the pulse parameters.