An effective method was established for the analysis of metabolomics in rice from different geographical origins by gas chromatography-quadrupole time-of-flight-mass spectrometry(GC/Q-TOF MS)，and was applied in geographic authenticity identification of rice.D-optimal experimental design was used to optimize the different extraction solvents，and the extraction efficiency was evaluated by partial least squares(PLS).Effects of derivatization reagents，derivatization temperature and time on the analysis of metabolites were investigated.The best result was obtained using methanol-water(1∶1) as extraction reagent，and derivatizing with N-methyl-N- (trimethylsilyl) trifluoroacetamide-trimethylchlorosilane(MSTFA-TMCS,99∶1) at 60℃ for 1h.10 highly matched metabolites were selected to assess the precision，reproducibility of the method and the stability of derived products.Satisfactory results were obtained with relative standard deviations(RSD) lower than 9%，confirming reliability of the method.Principal component analysis(PCA) prediction model was establised by mass profiler professional(MPP) to classify the Suijing 4 rice from four different geographical origins in Heilongjiang Province，and the results showed that samples from different places clustered into four different groups significantly.11 differential compounds were determined，which satisfied P<0.05 and a fold change cut-off ≥2.This demonstrates the feasibility of the developed analysis method in rice by GC/Q-TOF MS and its application in geographic authenticity identification of rice combined with PCA analysis.The method could provide a scientific basis for the standardization of rice market and the utilization of the resources.