Introduction: Lipidomics can offer an instant picture of the lipophilic metabolites from tissues and biofluids and can indicate the evolution of different pathologies, such as hyperplasia or different types of cancers. Related to these pathologies, Prostate Serum Antigen (PSA), proved to have a low grade prediction for an accurate diagnosis. Meanwhile, untargeted or targeted metabolomics became a useful advanced technology to discover new biomarkers for a better diagnosis.Aims: To realize an adequate procedure based on liquid chromatography coupled with mass spectrometry (HPLC-MS) to determine the profile of lipids from blood serum, followed by adequate biostatistics.Materials and Methods: Blood samples, obtained from healthy men and patients with prostate benign hyperplasia, post-biopsy cancer and post-surgery cancer were processed for lipid extraction and subjected to HPLC–ESI(+)QTOF-MS measurements, followed by the multivariate analysis (PCA and Cluster Analysis) with Unscrambler 10.1 software. TofControl 3.2 and Data Analysis 4.2 software (BrukerDaltonics) were used for the control of the instrument and data processing.Results: The molecules responsible for such discriminations were identified to be mainly represented by lyso-phospatidylcholines. By Cluster Analysis, the dendograms showed good statistical clustering of samples, especially for cancer patients agains controls and less clustered for hyperplasia.Conclusion: One can consider that molecules belonging to phospholipid family and diacyl /triacylglycerides or ceramides or carnitines can be considered potential biomarkers for hyperplasia and prostate cancer.