Abstract Purpose To analyze the plasma lipid spectrum between healthy control and patients with pancreatic cancer and to select differentially expressed tumor markers for early diagnosis. Methods In total, 20 patents were divided into case group and healthy control group according to surgical pathology. Of almost 1206 plasma lipid molecules harvested from 20 patients were measured by HILIC using the normal phase LC/MS. Heat map presented the relative levels of metabolites and lipids in the healthy control group and patients with pancreatic cancer. The PCA model was constructed to find out the difference in lipid metabolites. The principal components were drawn in a score plot and any clustering tendency could be observed. PLS-DA were performed to distinguish the healthy control group and pancreatic cancer according to the identified lipid profiling datasets. The volcano plot was used to visualize all variables with VIP>1 and presented the important variables with P<0.01 and -FC->2. Results The upregulated lipid metabolites in patients with pancreatic cancer contained 9 lipids; however, the downregulated lipid metabolites contained 79 lipids. Conclusion There were lipid metabolomic differences in patients with pancreatic cancer, which could serve as potential tumor markers for pancreatic cancer.