Extra virgin olive oil is favored by consumers for its high nutritional value，but its prices is relatively expensive．In order to increase profits，some illegal merchants have taken some concealed means to mix refined olive oil into extra virgin olive oil，or pass off refined olive oil as extra virgin olive oil to cheat consumers．At present，domestic studies on adulteration of olive oil mainly focus on the adulteration of olive oil with other edible oils，and few studies involve in grade identification of extra virgin olive oil and low-grade olive oil．Foreign studies on the identification of olive oil grades are mainly based on spectral technology，while the spectral method could only provide the relevant information of the whole fingerprint characteristics，but not the information of its specific components．Therefore，it is significant to establish a method to distinguish the grades of extra virgin olive oil and low-grade olive oils based on the characteristic components in olive oil．In this paper，a method was established to distinguish the grade of olive oil based on the content analysis of fatty acids combined with chemometrics．Extra virgin olive oil and refined olive oil with confirmed properties were used as test set．Fatty acids contents in the two grades of olive oil were determined by gas chromatography（GC） using SP－2560（100 m × 0.25 mm，0.2 μm） column．Olive oil grade discrimination model was established by three chemometrics，i.e. primary component analysis（PCA），hierarchical cluster analysis（HCA） and partial least squares－discrimination analysis（PLS－DA）．Results showed that，PCA could successfully distinguish extra virgin olive oil and refined olive oil，and HCA could also effectively identify two grades of olive oil．Through PLS－DA，six characteristic components with variable importance in the projection（VIP） greater than 1 were selected as follows：C23∶0，C18∶2n6t，C24∶0，C18∶1/C18∶2，C20∶1 and C18∶1n9c. Meanwhile，98 olive oil samples with unknown properties were taken as the validation set，and cross validation（CV） was carried out on the established olive oil grade discrimination model．The model prediction evaluation value（,Q,2,）,and correlation coefficient（,R,2,） were greater than 0.96．The results indicated that the established model of olive oil grade discrimination and prediction was reliable．Therefore，analysis of fatty acids and multiple element contents combined with chemometrics could be used to identify the grades of extra virgin olive oil and refined olive oil．