Liver cancer and healthy individual serum samples were compared based on their spectral features acquired by Raman and fluorescence spectroscopy to initially establish spectral features that can be considered spectral markers for liver cancer diagnosis. Intensity differences of the characteristic peaks of carotenes, proteins, and lipids in the Raman spectra were clearly observed in liver cancer patient serum samples compared to those of normal human serum samples. The changes in the serum fluorescence profiles of liver cancer patients were also analyzed. To probe the capacity and contrast of Raman spectroscopy as an analytical implement for the early diagnosis of liver cancer, principal component analysis was used to analyze the Raman spectra of liver cancer patients and healthy individuals. Furthermore, partial least squares-discriminant analysis was performed to compare the diagnostic performances of Raman spectroscopy for the classification of disease samples and healthy samples. Compared with existing diagnostic techniques, the Raman spectroscopy technique has many advantages such as extremely low sample requirements, ease of use, and ideal screening procedures. Thus, Raman spectroscopy has great potential for development as a powerful tool for distinguishing between healthy and liver cancer serum samples.