The use of Fourier transform infrared spectroscopy to identify the peritumoral tissue of gliomas proves the potential of this technique to distinguish normal brain tissues from glioma tissues. However, due to the heterogeneity of gliomas, it is difficult to characterize the representative spectra of normal brain tissues and glioma tissues. The linear spectra of major cellular components, such as microglia, astrocytes, and glioma cells, were obtained to quantify the biochemical changes between healthy cells and tumor cells, and provide supporting data for the final distinction between tumor and normal brain tissue. Fourier transform infrared was used to measure human astrocytes, microglia (HM1900), and glioma cells (U87, BT325), and the cellular components of the 4 types of cells were analyzed by means of average spectra, second-derivative spectra, principal component analysis, hierarchical cluster analysis, and difference spectra. The proteomics, lipidomics, genomics, and metabolic statuses of the cells were different. The amide I, lipid, and nuclear acid regions of the spectra are the most obvious regions to use for distinguishing the 4 types of cells. We conclude that an improved understanding of both similarities and differences in the cellular components of astrocytes, microglia, and glioma cells can help us better understand the heterogeneity of gliomas. We suggest that targeting cellular metabolism (protein, lipid, and nuclear acids) is helpful to distinguish between normal brain tissue and glioma tissue, which has broad application prospects.