The WRF (Weather Research and Forecasting Model) Version 3.2 was used to analyze the sensitivity of simulated airsea flux during a cyclone passage. Combinations of 9 microphysical parameterizations and 3 planetary boundary layer parameterizations were tested. LINYSU case has the strongest mean wind while WSM3ACM2 case has the weakest. Generally, the influences of planetary boundary layer on the simulation of sensible and latent heat flux are greater than those of microphysical schemes. Before the storm, all cases show that the airsea fluxes are not sensitive to the parameterization; during the storm, planetary boundary layer schemes influence the simulation of wind speed and air temperature, thus causing significant difference in simulated sensible heat fluxes; after the storm passage, test cases produce significant difference in both sensible and latent heat fluxes, which are mainly contributed by the combination effect of simulated wind speed, air temperature and specific humidity. Compared with observations, all simulated airsea fluxes have large systematic errors, which are probably due to the surface layer parameterization, indicating that the surface boundary layer parameterization in WRF 3.2 still needs to be improved.