Fluorescence spectra of sunflower oil under different oxidation conditions were obtained by three-dimensional synchronous fluorescence spectrometry, and the quality indexes were collected at the same time. The parallel-factor method was used to reduce the dimension of the three-dimensional synchronous fluorescence spectrum, and the mathematical models were established by iPLS, BiPLS and SiPLS pattern recognition methods. The results show: the two-dimensional synchronous fluorescence spectra of the samples have significant differences when the wavelength difference Δλ=50 nm, which are used for mathematical modeling of initial values. The results of partial least squares modeling show that the correlation coefficients of calibration set and prediction set of iPLS, BiPLS, SiPLS are 0.908 3, 0.961 2, 0.954 5 and 0.872 3, 0.925 2, 0.852 5. It’s found that BiPLS method is test. The study provides the theoretical basis and technical support for rapid identification of sunflower oil quality, and offers a theoretical basis for other related oil rapid detection.