a b s t r a c tThis paper describes the characterization of ciders (both ''natural" and sparkling cider) from the Principality of Asturias (northwest region of Spain) through the analysis of their protein content, based on their hydrophobic properties, and their foam characteristics. A reversed-phase high performance liquid chromatography (RP-HPLC) was applied to the protein analysis, and the foam parameters were measured with Bikerman's method. Multivariate techniques allowed the authors to differentiate ciders on the basis of the press and foam taking technologies, and foam sensory quality. Feasible and robust models were constructed for classifying purposes. Higher than 95% correct classifications were obtained for differentiating ciders on the basis of the factors studied (cider making technology and foam sensory quality). The multivariate regression model computed allowed the authors to predict (correlation coefficients higher than 0.8) the foam parameters related to foam stability and bubble average lifetime in ''natural" cider.