Due to its nutritional qualities, donkey milk is a newly popular food. Because it has a high added value, it may be adulterated with using cheaper milks, like cow milk. This study has investigated a rapid method for the authentication of pure donkey milk using Raman and FTIR-ATR spectroscopy and comparison of the methods. Three preprocessing methods were applied to the spectra. The results show that donkey milk has lower fat and protein content compared with cow milk. Notably, Raman spectroscopy successfully distinguishes donkey and cow milk according to the presence and absence of beta-carotene. Principal component analysis demonstrated a distinct separation between cow, adulterated donkey, and donkey. The variance value of 90.30% (PC1 = 72.76, PC2 = 17.53) is obtained from the first and second PCs for Raman data, and the variance value of 89.67% (PC1 = 65.25, PC2 = 24.41) is obtained from the first and second PCs for FTIR data. The Raman data could be used to separate donkey and cow milks, whereas the FTIR data were insufficient. It was observed that adulterated species could be separated between classes with Raman and FTIR. In the FTIR spectrum, there is a broad peak due to water, which accounts for about 87% of the milk composition, but this water peak is not included in the Raman spectrum. The results demonstrate the potential of Raman spectroscopy as a rapid and reliable method and suggest that it can be used as a nondestructive analytical tool for adulteration detection in donkey milk.