A rapid identification method of transgenic rapeseed oil based on near infrared spectroscopy was proposed. As transgenic food attracted more and more attention and in order to meet the requirement about labeling transgenic food, an accurate, fast, easy, efficient and low cost detection technology is needed. In this study, 117 rapeseed oil samples (including 64 non-transgenic rapeseed oil and 53 transgenic rapeseed oil samples) were used as test material. Principal component analysis (PCA) and discriminant partial least squares (DPLS) were applied to classify transgenic and non-transgenic rapeseed oil. Firstly, the paper studied the plot of principal component scores for the NIR spectra of transgenic and non- transgenic rapeseed oil. The result was that Non-transgenic and transgenic samples can be broadly distinguished in principal component space. Secondly, 35 samples were used to build DPLS identification model, and the other 82 samples as prediction set samples were identified by the model. The overall correct identification rate acquired was 96.34%, in which the correct identification rate of non-transgenic rapeseed oil was 95.56% and the one of transgenic rapeseed oil was 97.30%. The results showed that the attempt to discriminate transgenic rapeseed oil using NIR is feasible and excellent classification can be obtained by combining DPLS method.