Abstract: One-month data assimilation experiments were conducted to examine the impact of assimilating the intensive radiosonde observations in the East China regional numerical model system. Results indicated that assimilating intensive radiosonde observations had influence on geopotential height, specific humidity, temperature and wind speed at different heights in the model initial time. The influence on geopotential height and wind speed is relatively significant at high-level from 100 hPa to 150 hPa. The influence on specific humidity is relatively significant at low level from 700 hPa to 750 hPa. The model initial field obtained from including intensive radiosonde observations was more close to the observed results. The statistical results showed that assimilation of intensive radiosonde observations improves the forecast accuracy for heavy rainfall forecast and synoptic field forecast in some extent. The threat scores of 24 h rainstorm and heavy rain forecast were improved about 2.5% and 8.1%, respectively.