Based on the continuous observations on the land-atmosphere flux exchange and the meteorological and biological elements in a maize field at the Jinzhou Agricultural Ecosystem Research Station in Liaoning Province of Northeast China from 2006 to 2008, and by using the dynamic roughness (z0) and zero-displacement (d) parameterization scheme considering the effects of leaf area index, canopy height and wind speed at different developmental stages of maize, the BATS1e model was modified, and applied to investigate the effects of dynamic aerodynamic parameters on the flux exchange between maize agroecosystem and atmosphere. Compared with the original model, the drag coefficient (C(D)) simulated by the modified model increased, and its diurnal variation was more obvious with increasing vegetation coverage, which was more accordant with practical circumstances. The simulation accuracies of sensible heat (H), latent heat (lambdaE) and soil heat flux were improved in varying degree, and the Nash-Sutcliffes (NSs) were increased by 0.0569, 0.0194 and 0. 0384, with the improvement quantities in the growth season being 0.9%, 1.1% and 1.2% of global radiation, respectively. The dynamic parameterizations of z0 and d played a more remarkable role to increase the simulation accuracies of H and lambdaE with the actual observation of soil water content introduced into the improved model. This research proved that more reasonable dynamic aerodynamic parameterizations could fulfill an obvious function to improve the land surface processes simulation.