The formula of flavors is the core technology in the cigarette industry.Adding flavors to the cigarette during cigarette production can significantly improve the quality of the cigarette.However,traditional techniques used to identify the types of flavors in cigarettes are mainly through artificial smell and determination of certain physical properties such as refractive index,relative density,acid value,total volatiles,and chromatography-mass spectrometry.These techniques have many drawbacks,such as subjectiveness,complicated operation,time-consuming pretreatment,high cost,and low sensitivity.Based on this,this paper developed a method for rapidly identifying different types of tobacco flavors in production by surface-enhanced Raman spectroscopy with the assistance of principal component analysis.In this paper,the enhanced Raman spectra of different types,different batches,and different flavor additives were obtained.Principal component analysis showed that the scatter points of different types and different blending samples were distributed in different regions,while different batches of the same kind were concentrated in the scatter plot,which facilitating the differentiation.In this paper,surface-enhanced Raman spectroscopy with the assistance of the principal component analysis algorithm for the detection of tobacco flavors is simple,rapid,low-cost,and easy to analyze and identify.It is suitable for the quality control of tobacco flavors in the actual production.