In the paper 3 kinds of animal for a total of 30 animal blood samples and 21 human blood samples were chosen as the typical tested objects. Principle component analysis (PCA) method was employed to quickly identify the species of blood samples based on Raman spectroscopy. Vector normalization was used to preprocess the Raman spectroscopy signal. According to the plot of Leverage value vs. Studentized residue, outlier sample was detected and removed. By the PCA method, the correct rate between animal blood samples and human blood samples was up to 95% based on PC1and PC2. Further on the basis of PC4 and PC5, the correct rate of identification the animal blood sample was above 90%. The experiment results showed that the PCA method had good application prospects and feasibility to identify the species of blood samples. This method provided a reference for processing the similar problems in the field of forensic science and life science.