The demand of honey with defined quality has increased around the world; therefore, an adequate description of the traceability and authenticity of honeys is necessary. The pollen and physicochemical characteristics of 58 honey samples collected from five different environmental units (EUs) of the Jujuy province (Argentine) were determined, in order to differentiate them by geographical origin through the application of chemometric methods. A qualitative pollen analysis was performed by microscopy. The physicochemical characteristics were determined by Association of Official Analytical Chemists methods. Correspondence analysis (CA), principal component analysis (PCA) and linear discriminant analysis (LDA) were performed. Forty‐five per cent of honeys were monofloral; nearly half of them correspond to native species—among which, Salix humboldtiana, Baccharis sp. and Ziziphus mistol stood out. Physicochemical and microbiological analysis showed acceptable quality for honeys. CA characterized significantly honeys from four different EUs using all pollen data. While forward stepwise coupled to LDA identified 14 significant variables to build a discriminatory model with three significant discriminant functions, a cumulative variance of 94% was reached. Thus, 86% of the samples were correctly classified. PCA and LDA applied to physicochemical data allowed to distinguish three different groups with a significant function (p < 0.01) that explained 90% of the total variability. Honey corresponding to EU III Subandean Hills of Santa Bárbara was the most accurately classified by pollen and physicochemical data. The application of chemometric methods to pollen data and physicochemical parameters may be a useful tool to determine honey geographical origin. Copyright © 2014 John Wiley & Sons, Ltd.