Data assimilation plays an important role in numerical weather prediction. Radar radial velocity data with high spatial and temporal resolution are one of the most useful data in data assimilation. In order to investigate the effects of radar radial velocity data and data assimilation interval on typhoon Matsa’s track and precipitation, the direct assimilation procedure of Doppler radar radial velocity data using the three dimensional variational (3DVAR)data assimilation method is set up for the Weather Research and Forecasting (WRF) model. Five numerical experiments are conducted for the simulation of typhoon Matsa (2005): a control run (CTL) without data assimilation, and four assimilated runs (ASSG, ASSGR, ASS6h, and ASS3h). Comparing the assimilation experiments (ASSG and ASSGR) and the control one (CTL) shows that, the radar radial velocity data can tune the typhoon structure and hence improve the simulation of typhoon Matsa’s track and precipitation. During the first 12 hours, the deviation between simulation and observation of the Matsa position decreases by 19.3 km and 25.9 km for ASSG run and ASSGR run, respectively. Although the simulated 12 hour accumulated precipitations are both overestimated by the ASSG and ASSGR runs in Ningbo, Zhoushan, and Jiaxing regions, it is much better than that by the CTL run. In order to investigate the effects of data assimilation interval on typhoon Matsa’s track and precipitation, three assimilated experiments with different data assimilation intervals are compared. Results show that small data assimilation interval can improve the result of the numerical experiment.