Mid-infrared regions of the spectrum of grain of four Triticum species were analyzed using Fourier transform infrared – attenuated total reflection. Significant variations were noted in the absorbance of all studied taxa over four wavenumber ranges. The principal component analysis supported strong discrimination of the four examined species. The percentage of variation explained by the first two principal component analyses reached 95.04%, including principal components 1-72.16% and 2-22.88%. The applied method supports quick identification of the grains of various hulled species of wheat and it is a useful tool for evaluating the seeds and food products obtained from those cereal species.