The aim of this study was to find lipid metabolite concentrations differences between healthy subjects and colorectal cancer patients. A novel high performance liquid chromatography–electrospray ionization tandem mass spectrometry (HPLC–ESI–MS/MS) analytical method was developed for the analysis of nine lipid series (two sphingolipids, seven phospholipids) using the multiple reaction monitoring mode (MRM) on dried plasma spots. The extracted lipids were separated by HPLC using an Imtakt Unison UK-C8. The developed method was applied to human plasma samples obtained from healthy subjects (n = 40) and colorectal cancer patients (n = 40). A partial least squares discriminant analysis (PLS-DA) model, which is a multivariate statistical analysis method, was employed to analyse the quantitative results. The VIP score of the PLS-DA model was employed to effectively discriminate colorectal cancer patients from healthy subjects. Furthermore, a random forest classification method was employed. Through statistical processing, the lipid 18:1 Lyso PC in accordance with VIP score > 1 was identified, and four lipids in accordance with p-value < 0.05, 15:0–18:1 PC, 18:1 Lyso PC, 18:1 Lyso PE, and C15 Ceramide (d18:1/15:0) were identified. The results of this study corresponded to study of Zhao et al.(2007) that 18:1 LPC can be potential biomarker between normal and colorectal cancer patients. This method is expected to be practically useful due to its simple dried plasma spot use and excellent sensitivity for lipid screening when applied to various diseases, including colorectal cancer.