East Asia is one of the most sensitive and vulnerable areas in the world to the global climate changes. The study of the climate changes in this region is much more difficult than other places. With high quality, long time series and high resolution, atmosphere reanalysis has become the most widely used datasets in atmospheric science. However, the systematic biases can influence the quantity of reanalysis datasets. Therefore,an evaluation of the reliability and accuracy of reanalysis datasets is significant for global and regional climate research. In recent years,lots of work have been made to investigate the reliability of reanalysis. However, most of them focused on surface variables, few reanalysis and the time period were limited. This work evaluated the upper-level variables extracted from following sources : the National Centers for Environment Prediction (NCEP)-National Centers for Atmospheric Research (NCAR) reanalysis-1,the NCEF-Department of Energy (DOE) renalysis-2, the NCEP-Climate Forecast System Reanalysis (CFSR),the 25-year Japanese Meteorological Agency (JRA-25) reanalysis, the European Centre for Medium-Range Weather Forecast (ECMWF) Interim reanalysis (ERA-Interim) and the Modern-Era Retrospective analysis for Research and Applications (MERRA) reanalysis products, and Integrated Global Radiosonde Archive (IGRA) global sounding observations over China from 1989 to 2008. Three sets of reliability and accuracy studies were carried out: the first evaluated the spatial distributions of summer mean values of the several high-variables represented by all reanalysis mentioned above. The second aimed to assess the performance of the inter-annual variation of high-variables described by reanalysis. The third compared the similarities of empirical orthogonal function (EOF) modes between observations and reanalysi s. It was found that the mean values of geopotential height and temperature in each reanalysis dataset are consistent with the observations, but the wind fields, especially the meridional wind,are not. Besides,the reanalysis products do a bad job in revealing the interannual variation of meridional wind. The results of EOF analysis imply that all reanalysis datasets exhibit better performance in depicting the temporal and spatial distributions of geopotential height and temperature than other variables,especially the wind fields;MERRA performs specific humidity better than other reanalysis products. Generally, NCEP/NCAR, NCEP/DOE and NCEP/CFSR products are not as good as JRA-25, ERA-Interim and MERRA. Based on the study, we recognized the differences of each reanalysis in describing the characteristics of upper-level variables. In this way, more desirable alternative will be made when considering which reanalysis will be used in climate change research. However, most of these reanalysis do a bad work in revealing the wind field, which may be caused by the wind rotation. Hence, more studies are needed on the ability of reanalysis in representing wind field in the future. Besides,only upper-level variables have been compared,the research on surface variables,such as surface temperature and precipitation, can be added in the next step. Nowadays, tie field of regional climate modeling is booming due to enormous demands for prediction of future regional climate change, downscaling seasonal prediction, and for use in a variety of regional climate applications. It is, therefore,imperative to better understand the strengths, deficiencies, and limitations as well as the sources of uncertainties that can occur with regional climate model (RCM) simulations. As we know,running a RCM needs lateral boundary (LB) forcing fields and initial conditions. Hence, different LB forcing fields and initial conditions may cause different simulation results. In the future,the uncertainties of RCMs caused by different driv ng reanalysis datasets will be researched.