The second derivative transformation and loadings plot over the range of 1800–900 cm−1 highlighted the most distinctive variations among bacterial spectra. Loadings 1 and 2 were distinctively representative of the bacterial spectral data and accounted for 73% of the total variability. Treatments were noticeably segregated with distinct clustering by principal component analysis (PCA). Using soft independent modeling of class analogy (SIMCA) analysis, 88.3% of (E. coli O157:H7 ATCC 35150) spectra, 75.0% of (A. acidoterrestris 1016) spectra, 88.3% of (Alicyclobacillus spp. C-Fuji-6) spectra, and 80.0% of the mixed culture of both Alicyclobacillus strains spectra were correctly classified. Using the spectral features of bacterial cellular constituents such as nucleic acids, proteins, phospholipids, peptidoglycan, and lipopolysaccharides from examined bacterial cells, pure and mixed cultures of Alicyclobacillus spp. cells, and the pathogenic E. coli cells could be detected in apple juice.