The databases correspond to research data obtained in the doctoral thesis entitled: Near-infrared spectroscopy and chemometrics for the authentication of Brazilian artisanal cheeses. The first data file compiles information used to develop a non-destructive method, based on near-infrared (NIR) spectroscopy, for identifying adulteration in butteroil (Manteiga de garrafa) resulting from the partial replacement of milk fat with lard, soybean oil, or expired butter. It contains physicochemical results (chemical composition, fatty acid profile, acid value and peroxide value) and instrumental color parameters for authentic samples, as well as NIR spectra of authentic butteroil, adulterated samples, and the respective adulterants. The second date file compiles information used to develop a non-destructive method, based on near-infrared (NIR) spectroscopy, to identify adulteration in queijo de manteiga (Brazilian butter cheese) by replacing manteiga de garrafa (butteroil), its main ingredient, with soybean oil. The file contains physicochemical results (chemical composition and fatty acid profile), instrumental color parameters, and NIR spectra (900–1700 nm) of authentic and adulterated queijo de manteiga samples. And the third data file contains information used to discriminate varieties of Brazilian artisanal cheeses and to predict chemical composition employing non-destructive technologies: a portable NIR spectrometer (900–1700 nm) and hyperspectral imaging in the vis–NIR (397–1004 nm) and NIR (884–1717 nm) ranges. Parameters of chemical composition, fatty acides profile and instrumental color were assessed to capture inter-cheese variation. These analyses support the development of classification models and predictive models to support artisanal origin certifications.