A new method for identification of adulterated sesame oil is proposed, and the method combines three-dimensional fluorescence spectroscopy with wavelet compression and alternating penalty trilinear decomposition (APTLD) algorithms. The three-dimensional fluorescence spectra of pure sesame oil and adulterated sesame oil samples are measured with fluorescence spectrometer, and the error caused by the instrument is eliminated via excitation correction and emission correction, then the true three-dimensional fluorescence spectrum data of the samples are obtained. Wavelet compression is used to compress the processed real data to reduce redundant information. The compression fraction and data recovery fraction are greater than 94.00% and 98.00%, respectively. The APTLD algorithm is used to qualitatively and quantitatively analyze the compressed three-dimensional fluorescence spectral data. The obtained recovery rate is 97.0%-99.8%, and the root mean square error of prediction (RMSEP) is not larger than 0.120. The research results show that the proposed method can accurately identify the pure sesame oil and the adulterated sesame oil samples, and predict the component of the samples.