As the problem of product adulteration has always existed in the ghee industry, there is an urgent need exists for a reliable and rapid method to distinguish the authenticity of ghee. In this study, adulterated ghee samples with different concentrations were identified by high-resolution Raman spectroscopy combined with multivariate data analysis. The band differences of 1007, 1156, and 1520 cm−1 assigned to β-carotene and 1656 cm−1 assigned to CLA in the Raman spectra of ghee were investigated. The partial least squares discrimination analysis (PLS-DA) model established in the whole spectral range was able to identify the yak ghee samples adulterated with cattle-yak ghee of more than 5% mass fraction, and the total correctness of both the training and prediction sets was 100%. The results demonstrated that Raman spectroscopy combined with chemometric methods has a broad application potential in the identification of adulterated ghee products as a functional analysis technique for rapid detection. •Raman spectroscopy was used to characterize the spectra of Yak ghee.•The adulteration model of ghee using partial least squares discriminant analysis.•This method is a rapid and less labor-intensive way to identify adulterated ghee.