Mixing of food commodities with low-quality substances are being widely practiced around the globe that impose negative effects on the consumers’ health. Addition of different vegetable oils in butter is a common practice that deteriorates the quality of butter. Suppliers are partially or completely replacing the butter fat with different low-quality vegetable oils and animal fats. Currently available methods used for detection of fat adulteration in different dairy products are color test and apparent solidification test. These methods are no doubt reliable but not comprehensive, require expensive and hazardous chemical reagents and are time consuming. Hence there is need to use some rapid and non-destructive methods which are suitable and can serve the purpose. Present study was aimed to detect the adulteration of vegetable oil in butter with FTIR spectroscopy. Butter samples with different concentrations of vegetable oil (10, 20, 30, 40, 50, and 60%) were prepared and FTIR spectra were collected (4000–650 cm⁻¹). PCA were applied on data set of FTIR spectra recorded on butter, vegetable oil and adulterated samples after removing the outliers using H-statistics. The PCA similarities maps defined by the principal components 1 and 2 for FTIR spectra showed that the first 2 principal components accounted for 98% of the total variance with a predominance of component 1 (85%). Better discrimination of butter, vegetable oil and adulterated samples were observed after removing outlier from the original data set. These results confirmed that the FTIR spectra retained information related to the molecular structure of butter and vegetable oil allowing discrimination of samples as a function of adulteration. The results obtained by PLS (partial least squares) regression model showed the values of R² equal to 0.88 for calibration and 0.68 for validation using 3 PLS factors. The good predictions strongly suggest that the FTIR spectra retain information on butter adulteration and that PLS regression can extract specific information regarding spectral variations due to the addition of vegetable oil in butter samples, which are helpful in predicting the adulteration in the butter samples.