Abstract. Nearly one quarter of patients with colorectal carcinoma (CRC) were diagnosed at an advanced stage. Under these circumstances, radical resection of the tumor is the best strategy to enhance the five-year survival rate. However, up to 50% of post-operative patients experience cancer recurrence within the first few years. Therefore, post-operative surveillance is important. However, currently performed post-operative monitoring relies on relatively dated methods with insufficient sensitivity and specificity. The present study applied an advanced technology of ultra-performance liquid chromatography coupled with quadrupole time-of-flight mass spectrometry in order to examine changes in metabolite patterns in serum with the aim of identifying reliable biomarkers in patients with CRC at various time-points. Serum samples were collected from and 20 CRC patients prior to radical resection (group 1) and one month following radical resection (group 2) as well as from 20 healthy volunteers (group 3). Multivariate pattern recognition was used to identify potential biomarkers of CRC. Compared with healthy volunteers, three groups of biomarkers were identified in patients with CRC (P<0.05), namely phosphatidylcholines (PCs), lysophosphatidylcholines (LPCs) and diacylglycerols (DAGs). However, no statistical difference in the levels of these biomarkers between pre-operative and post-operative CRC patients was identified (P>0.05). PCs and LPCs, which contain polyunsaturated fatty acids, were decreased, whereas LPCs and DAGs, which contain saturated fatty acids, were increased in CRC patients. The present study demonstrated that obvious metabolic disturbances occur during the development of CRC and provided a novel analytic method, which is likely to be used as a diagnostic tool for CRC and may help to improve the patients' prognosis.