A multisensor system combined with multivariate analysis is applied for the characterization and classification of white grape juices. The proposed system, known as hybrid electronic tongue, consists of an array of electrochemical microsensors and a colorimetric optofluidic system. A total of 25 white grape juices representing the large variability of vines grown in the North-west Iberian Peninsula were studied. The data obtained were treated with Principal Component Analysis (PCA) and Soft Independent Modeling Class Analogy (SIMCA). The first tool was used to train the system with the reference genotypes -Albariño, Muscat à Petit Grains Blanc and Palomino- and the second to study the feasibility of the hybrid electronic tongue to distinguish between different grape juice varieties. The results show that the three reference genotypes are well differentiated in the PCA model and this can be used to interpolate the rest of varieties and predict their basic characteristics. Besides, using the SIMCA, the system demonstrates high potential for classifying and discriminating grape varieties.