NMR and electronic tongue analysis coupled with multivariate statistical data was used to characterize the Cabernet Sauvignon and Merlot dry red wines from Ningxia and Hebei vinified in 2010 vintage. The purpose of this study was to compare the characterization of dry red wines from two characteristic regions and analyze the difference of their metabolites. Results from 1H NMR analysis coupled with pattern recognition methods showed that the Cabernet Sauvignon and Merlot dry red wines from two different regions could be separated. Compared with dry wines from Hebei,Cabernet Sauvignon dry red wines from Ningxia contained higher levels of valine,succinic acid and glycerol,whereas 2,3-butanediol,proline,lactic acid,acetic acid,ethyl acetate,choline,gallic acid were in lower levels. Higher levels of Valine,proline,succinic acid,acetic acid,and glycerol were found in Merlot dry red wines from Ningxia,whereas the levels of Lactic acid,2,3-butanediol,choline,gallic acid,ethyl acetate were lower according to the loading plot of PLS-DA. All data from the electronic tongue analysis were treated by multivariate data processing based on PCA,and the result showed that the electronic tongue was able to identify the dry red wines from different regions. This study using two analysis methods showed that there were significant differences between Cabernet Sauvignon and Merlot dry red wines from Ningxia and Hebei.