Abstract: In order to find out a fast quantitative determination method for the essential fatty acids in edible oils, near infrared spectroscopy was applied in determination of the contents of α-linolenic acid and linoleic acid. The chemometrics models between near infrared spectra and the contents of α-linolenic acid and linoleic acid were established by partial least squares regression (PLS) and least-squares support vector machine (LS-SVM). Several common used spectral pretreatment methods were used to establish different PLS and LS-SVM models. For linoleic acid prediction, best predictive performance was obtained using LS-SVM model and spectral pretreatment of Savitzky-Golay smoothing and multiplicative scatter correction (MSC). Coefficient of determination (Rp2), root mean square error for prediction (RMSEP) and residual predictive deviation (RPD) were 0.0161, 0.989 and 9.4783, respectively. For α-linolenic acid prediction, best predictive performance was obtained using LS-SVM model and the spectral pretreatment of Savitzky-Golay smoothing and standard normal variation (SNV). Rp2, RMSEP and RPD were 0.0036, 0.972 and 6.0561, respectively. The results indicate that it is feasible to use near infrared spectroscopy for fast determination of α-linolenic acid and linoleic acid contents in edible oils.