Histamine is a kind of biogenic amine， which is produced by decarboxylation of histamine acid. It is one of the by-products of Huangjiu fermentation and is closely related to human health. In this study， surface-enhanced Raman spectroscopy (SERS) and chemometrics were combined to detect histamine in Huangjiu quickly. SERS spectra of histamine were obtained by using silver nanoparticles (Ag NPs) as SERS substrate and NaCl solution as aggregating agent. Firstly， the SERS reaction conditions were optimized by using the concentration of histamine aqueous solution to NaCl solution， the mixing volume ratio of histamine to NaCl and Ag NPs， and the mixing time of the three. Then， based on the optimal conditions， the purification effects of C18 solid-phase extraction column and PXC strong cation exchange reversed-phase solid-phase extraction column on Huangjiu were further compared to quantitatively analyze histamine in Huangjiu. The results showed that when NaCl solution concentration 0.5 mol/L， histamine solution∶NaCl solution∶Ag NPs=2∶1∶3， and mixing time 15 s， the enhancement effect at the characteristic peak of histamine at 1 002 cm-1 was the strongest. Under this condition， the detection limit of histamine solution was 1 mg/L. At the same time， under this condition， PXC strong cation exchange reversed-phase solid phase extraction column was determined as the pretreatment method of Huangjiu， and the detection limit of histamine in Huangjiu was 2 mg/L. Finally， multiple linear regression analysis and partial least squares regression (PLSR) prediction models of histamine content are established， and the effectiveness of competitive adaptive weighted sampling (CARS)， Principal component analysis (PCA) and Successive projections algorithm (SPA) in screening characteristic variables of original spectrum was compared. Among them， the calibration set and verification set of PLSR model established by using 51 variables screened by CARS have obtained higher determination coefficients Rc2=0.9777， Rv2=0.9604， and lower root mean square error RMSEC=0.0712 mg/L and RMSEV=0.1405 mg/L. The results showed that the rapid quantitative detection of histamine in yellow rice wine could be realized based on SERS technology and chemometrics method， which provided a new idea for the on-site monitoring of histamine in fermented alcoholic beverage.