Current novel drug developments for the treatment of diabetes require multiple bioanalytical assays to interrogate the cellular metabolism, which are costly, laborious and time-consuming. Fourier-transform infrared (FTIR) spectroscopy is a nondestructive, label-free, sensitive and low-cost technique that is recently found to be suitable for studying living cells. The aim of this study is to demonstrate that live-cell FTIR can be applied to study the differences in glucose metabolism in cells in normal culturing medium and cells treated in high glucose (a diabetes model) in order to highlight the potential of the technique in diabetes research. Live HepG2 cells were treated in normal glucose (3.8 mM; control) or high glucose (25 mM) medium and were measured directly using the FTIR approach. Principal component analysis was used to highlight any possible correlated changes 24, 48 and 72 h after treatments. FTIR spectra of live cell treated in normal and high glucose medium have shown significant differences (p < 0.05) for all treatment time. The control cells have seen an increased in the absorbance at 1088, 1240 and 1400 cm−1, which are associated with phosphate stretching mode vibrations from phosphorylated proteins and DNA back bone; and symmetric stretching mode vibration of COO- from fatty acids, amino acids, lipids and carbohydrate metabolites. However, the high glucose treated cells have shown a different changes in the 1000–1200 cm−1 region, which is linked to the glycogen and ATP:ADP ratio. In conclusion, live-cell FTIR can be a low-cost method for the studies of metabolic changes in cells.