Rationale Development of therapy-resistant cancer is a major problem in clinical oncology, and there is an urgent need for novel markers identifying development of the resistant phenotype. Lipidomics represents a promising approach to discriminate lipid profiles of malignant phenotype cells. Alterations in phospholipid distribution or chemical composition have been reported in various pathologies including cancer. Here we were curious whether quantitative differences in phospholipid composition between cisplatin-resistant and -sensitive model cancer cell lines could be revealed by mass spectrometric means. Methods The phospholipid contents of cell membranes of the cancer cell lines CCRF-CEM and A2780, both responsive and resistant to cisplatin, were analyzed by solid-phase extraction (SPE) and electrospray ionization mass spectrometry (ESI-MS and tandem mass spectrometry (MS/MS)). Extracts were obtained by disruption of cells with a dounce tissue grinder set followed by centrifugation. To minimize the enzymatic activity, phospholipids were extracted from cell extracts by SPE immediately after the cell lysis and analyzed by MS. Both supernatant and pellet fractions of cell extracts were analyzed. Results A phospholipid profile specific for cell lines and their phenotypes was revealed. We have documented by quantitative analysis that phosphocholines PC P-34:0, PC 34:1, PC 20:2_16:0, LPC 18:1 and LPC 16:0 PLs were present in the 200-400 mu M concentration range in CCRF-CEM cisplatin-responsive cells, but absent in their cisplatin-resistant cells. Similarly, PC 34:1, LPC 18:1 and LPC 16:0 were increased in cisplatin-responsive A2780 cells, and PC 20:2_16:0 was downregulated in cisplatin-resistant A2780 cells. Conclusions In this work we showed that the ESI-MS analysis of the lipid content of the therapy-resistant and -sensitive cells can clearly distinguish the phenotypic pattern and determine the potential tumor response to cytotoxic therapy. Lipid entities revealed by mass spectrometry and associated with development of therapy resistance can thus support molecular diagnosis and provide a potential complementary cancer biomarker.