Staphylococcus aureus is an extremely infectious and malignant pathogen among many bacteria species. The aim of this work is to provide a robust classification model that would be able to identify S. aureus independent of the culture growth stage and the variations in bacteria concentration in suspension and also one that would be able to identify the pathogen among both taxonomically close species of the same genus and taxonomically distant species of different genera, using Fourier transform infrared spectroscopy (FTIR). In total, the spectra of 141 isolates of 17 bacteria have been used. Based on a combination of principal component analysis (PCA) and linear discriminant analysis (LDA), an identification model providing 100% sensitivity and 98% specificity was built. Inherent reliability and flexibility of the model have been shown. The proposed method of analysis allows us to get closer to the diagnostic requirements in the field of clinical microbiology, and it can be utilized for typing of other pathogenic bacteria species.