As a non-destructive analysis technology，Raman spectroscopy, has been widely used in writing materials analysis in the field of document examination in recent years. The main method to identify the inks by Raman spectrum is direct comparison of the number, the position and the feature of peak. Chemometrics is the discipline that creates a link between measured values and the system state in chemical system through statistical or mathematical method, which can be used for the analysis of spectral data. This study was combined the similarity analysis and principal component analysis (PCA) in chemometrics with Raman spectra data of 28 blue gel pen inks collected on market. The results of classification by two analysis methods were almost accordant, which the 28 blue gel pen inks could be divided into five categories using two methods. The results showed that the data obtained from Raman spectroscopy combined with chemometrics analysis method for data analysis and identification, was more objective and accurate than the direct comparison of spectra.