Identification of NMR-based metabolic indexes is limited by the deleterious effects of copious proteins and lipoproteins in the serum that accentuate the need for advance and high-throughput method. We tried to explore the use of a novel filtration (2KDa molecular weight cut-off) approach to remove the proteins from serum following use of less sample volume (only 150 μL of filtered serum), combining an array of 1D/2D NMR experiments (at 800 MHz spectrometer), spiking experiments with standard compounds, and validated by mass spectrometry. This novel method enabled the identification of a large number (n = 73) of metabolites and their percentage of abundance in the present study cohort. Mass spectrometry further validates and confirms the presence of all these 73 metabolites using same filtered serum. This study reveals seven new metabolites (citrulline, inosine, taurine, trimethyl amine, methylmalonate, uracil, methanol) in filtered serum using 1D/2D NMR spectroscopy that were not observed in earlier available literature using protein precipitation approach. This novel method delineates volatile metabolites, nitrogenous bases and nucleosides that may provide a milestone for the identification of inborn error of metabolism, pathogenicity at molecular level, disease identification and prognosis, and forensic studies using minimal volume of filtered serum samples and NMR spectroscopy.