Exosomes are nano-sized phospholipid bilayer-enclosed vesicles secreted by all cells into the extracellular milieu. Released exosomes contain cell-specific proteins, membrane lipids, mRNA, DNA and microRNA that can perform versatile roles in normal or diseased processes. Exosomes are ideal biomarkers of cancer, which have important application potential in liquid biopsy and are expected to become one of the means for rapid cancer detection. Surface enhanced Raman spectroscopy (SERS) is a molecular vibration spectrum, which can detect the fine structure and information changes of substances at the molecular level, and has the characteristics of a “fingerprint spectrum”. In this study, the exosomes were isolated via differential centrifugation combined with ultracentrifugation of the supernatants of breast cells. The SERS profiles of breast cancer cells and their exosomes were collected with colloidal Au nanoparticles as the enhanced substrate, and multivariate statistical analysis was used to identify and distinguish breast cancer cells. The results showed that breast cancer cells and their exosomes had characteristic Raman signals in the range of 500～1 600 cm-1. Their Raman phenotypes obtained by non-labeled detection were the presentation of all the signals of the“whole-organism fingerprint” of the sample. The accuracy rate reached 100% by using exosome-SERS detection and OPLS-DA analysis. Single-cellular SERS detection combined with PCA-LDA analysis showed that the accuracy of differentiating breast cancer cells was 83.7%. Breast cancer cells and their exosomes showed similarity at the bands of 506～569 and 1 010～1 070 cm-1, but the characteristic Raman peaks of exosomes at 735, 963 and 1 318 cm-1 were significantly higher than those of cells. It may be because the structure of exosomes is simpler than that of cells, and the information of biological macromolecules such as nucleic acids and proteins can be characterized more easily, indicating the feasibility of rapid identification of breast cancer by detecting exosomes by SERS technology. In summary, this study established a non-labeling and direct detection method for rapidly detecting single cells and their exosomes by SERS analysis. Combined with multivariate statistical analysis, different types of breast cancer cells could be quickly identified, and the relationship between exosomes and maternal cells was explored from the perspective of Raman omics. This method has the advantages of non-labeling, rapidness, sensitivity, accuracy and simplicity, which provides an effective technical means for rapid diagnosis and screening of breast cancer in vitro, and lays a foundation for clinical application.