In order to evaluate the improvement of the GPS/PWV data assimilation to the initial humidity field and precipitation forecast field of numerical model, a series of assimilation experiments were performed using the LAPS (Local Analysis and Prediction System) and WRF (Weather Research and Forecasting) model. One heavy rainfall process in Hubei Area was taken as the experimental research object. The result shows that the assimilation of GPS/PWV is effective in reducing the error of PW in the LAPS humidity field, however its impact on potential height and wind fields is not so notable. The impacts of GPS/PWV are greater than radar data by comparing the total precipitable water analyzed by LAPS in different schemes. Different weather forecast scores show that the initial moisture field in WRF is significantly improved by using GPS/PWV data, and the improved moisture field in initial condition leads to positive effect on the forecast of rainfall.