The adulteration of edible oil seriously threatens consumers’ physical health and disrupts the social market order. Therefore, developing effective methods for identifying adulterated edible oil is crucial to establishing a safe and reliable food supply chain and enhancing consumer welfare. This article studies a method to identify adulterated edible oil using sesame oil as a case study. The study first formulated three types of contaminated sesame oil by adding sesame flavor, corn oil, soybean oil, and rapeseed oil. The FLS920 steady-state fluorescence spectrometer was then employed to collect 3D fluorescence spectrum data from 45 experimental samples, including these three types of contaminated sesame oil and different brands of pure sesame oil. Subsequently, two-dimensional features were extracted from the experimental samples using the 2D-LDA method. The principle of nearest-neighbor classification was applied to identify adulterated edible oils accurately. Moreover, the proposed method was compared with the PARAFAC-QDA and NPLS-DA methods. The results demonstrated that the 2D-LDA method effectively extracted two-dimensional features characterizing adulterated sesame oil. These features facilitated maximum separation of different classes of experimental samples in the projection subspace. Simultaneously, they allowed experimental samples of the same class to cluster closely in the subspace. The distinct characteristics of these features enhanced sample separability in the low-dimensional subspace, resulting in 100% identification accuracy. In contrast, the PARAFAC-QDA and NPLS-DA methods achieved 85% and 95% discrimination accuracies, respectively. Hence, the 2D-LDA method outperformed these two methods in identifying edible oil adulteration, offering a simpler and more accurate identification process and results. This study provides an efficient and feasible new solution for on-site food safety supervision.