To investigate the application and impact of Lband high vertical resolution minutelevel sounding data which include balloons accurate drifting information at each level on data assimilation and numerical prediction, basic features of these data were preliminarily analyzed in this paper. On this basis, experiment of single station, precipitation case study and batch assimilation tests of the whole simulated domain were conducted using minutelevel sounding data via the two assimilating methods of profile and pointbypoint, then compared with the assimilation test using conventional sounding data. It was concluded that: (1) the vertical resolution of Lband minutely sounding data can better match the vertical layers in the model. It can provide additional vertical observation information and precise location data that the conventional sounding data cannot achieve. (2) Onestation assimilation test indicates that assimilating minutely sounding data can efficiently increase the weight of observations in the analysis, thus leading to the analysis field more similar to the observations, and differences of analysis field calculated by pointbypoint assimilation method and profile way mainly remain in the upper troposphere. (3) The case study and batch tests assimilating the minutely sounding data in the whole forecast region can affect the final analysis and forecast accuracy, and the difference of wind assimilation caused by the balloon’s advection drift mainly exists in the positions at 200 hPa, and the upperlevel jet stream over northern China. (4) The initial field in the model is overall improved by assimilating Lband minutly sounding data, which mainly positively affect the forecast of winds in the upper troposphere and intense precipitation in larger scales.