This study evaluates the potential use of the regional climate model version 3 (RegCM3) driven by (1) the National Centers for Environmental Prediction (NCEP) Climate Forecast System Reanalysis (CFSR) data during 1982–2001 and (2) the NCEP Climate Forecast System Version 2 (CFS2) hindcast data during 2000–2010 in forecasting Western North Pacific (WNP) tropical cyclone (TC) activity. The first experiment is conducted to investigate the ability of the model in generating a good climatology of TC activity in spatial and temporal scales, so the model could be used in the second experiment to test its ability in forecasting TC genesis and landfall. Both experiments extend through the May to October WNP-TC season. Results show that the use of RegCM3 driven by the CFSR achieves a better simulation on the temporal and spatial variation of WNP-TC genesis during 1982–2001, as compared to previous studies using the same model but driven by the ERA40 reanalysis. In addition, diagnoses on the use of RegCM3 driven by the CFS2 point out that the 2000–2010 WNP-TC genesis locations and numbers from the model are very similar to those from the observations. The skill of RegCM3 in the forecasts of landfalling TCs is higher over the Southeast Asian region than over the other sub-regions of East Asia. Potential causes for such regional differences are discussed. Most importantly, statistical analyses show that the use of RegCM3 driven by the CFS2 gives a better forecast skill than the use of CFS2 alone for the prediction of WNP-TCs making landfall in East Asia. This indicates that the use of a dynamical downscaling method for the global forecast data would likely lead to a higher forecast skill of regional TC landfalls in most of the East Asian region.