Abstract: The wind loading is one of the crucial loads in both the wind-resistant design and disaster prevention and reduction analyses for overpass bridges. Inasmuch as the wind-overpass bridge interaction is rather complicated, we can obtain the reliable data of wind velocity (loading) by resorting to wind tunnel tests, full-scale measurements, and numerical simulations. Notwithstanding this, we may gain an insight into the dynamic characteristics of vibrations of overpass bridges under wind loads via the time history analysis. The effectiveness of wind vibration control of overpass bridges can be intuitively displayed in the time domain. Therefore, the main objective of this paper is to simulate the fluctuating wind velocity time series of overpass bridges with resorting to the auto-regressive move average (ARMA) model in the white noise filtration method (WNFM). First, taking into account both the time and space correlativity of the wind speed of overpass bridges, the expressions of the ARMA model have been derived with the order of both the auto-regressive (AR) and move-regressive (MA) components being unequal. Then, based on Kaimal spectrum, the simulation of the wind velocity time histories of a real overpass bridge has been carried out utilizing the ARMA model. Eventually, through comparing the simulated values of wind speed spectral density, auto- and cross-correlation functions with the corresponding targets, the flexibility is then corroborated of simulating the fluctuating wind speed time series of overpass bridges using the ARMA model.