(1)H NMR spectroscopy was applied to discriminate triple concentrated tomato paste coming from two different countries. Notwithstanding different tomato cultivars and ripening stages employed to obtain the final product, significant discrimination between Italian and Chinese samples was obtained by combining NMR data and principal component analysis. Supervised orthogonal projection to latent structure discriminant analysis (OPLS-DA) technique was used to build robust classification models, while S-plot was employed to identify statistically significant variables responsible for class separation. Citrate content resulted in being the most relevant chemical compound for Chinese and Italian sample differentiation. In order to highlight other compounds able to contribute to sample differentiation, citrate content was excluded, and a new classification model was built. This latter model indicated aspartate, glutamine, and sugars as important variables in sample differentiation.