In past years, NAEFS (North American Ensemble Forecasting System) high resolution probabilistic products have been applied to CONUS and Alaska region. The variables have been expended from surface pressure, temperature, winds to maximum/minimum temperature and wind speed/direction. In this study, the relative humidity and dew point temperature have been tested for 1) statistical bias correction at 1*1 degree globally for both NCEP GEFS and CMC GEFS; 2) statistical down-scale to NDGD (National Digit Guidance Database) resolution for CONUS and Alaska region. The post process for relative humidity and dew point temperature are different from other variables such as surface pressure and temperature. Relative humidity is bounded by two ends (0% and 100%). Dew point temperature is bounded by one end (less than temperature). There is no simple method to treat these bounded variables from statistical post process currently. However, it is possible to reduce this difficulty by calibrate dew point temperature first, then deriving relative humidity from post processed dew point temperature and temperature. In the future, a hybrid post process system may be considered through combining statistical and dynamical downscaling in order to avoid this problem.