The objective of this study was to investigate honey samples from two bee tribes grouped in Meliponini (stingless bee) and Apini (Apis mellifera) honeys from Brazil, in order to compare their chemical composition. Thirty-four honey samples were collected and evaluated by proton NMR spectroscopy, HPLC and physicochemical parameters, using chemometric methods. Principal component analysis (PCA) was able to discriminate Apini and Meliponini goups. Additionally, the partial least squares discriminant analysis (PLS-DA) models were able to predict the Apini and Meliponini honey samples with sensitivity, specificity and accuracy values close to 100%. Frequency values between 4.00 − 3.00 ppm, corresponding to the sugar region, were important for the classification by the PLS-DA model using NMR data. For the PLS-DA model using physicochemical data, the main parameters for classification were water activity, reducing sugar, moisture, refractive index, total soluble solids, maltose and glucose. Since stingless honeys have no regulation, the data presented in this study can compose a database to assist in the establishment of regulatory criteria. Moreover, the chemometric models proposed can be used as a tool for the quality control of honey, especially for authentication purposes.