A novel approach for metabolite extraction and fingerprinting was introduced based upon the nano-baskets and emulsion liquid membrane-nuclear magnetic resonance (ELM-NMR) technique. The objective of this method is optimizing the fingerprints, minimizing the metabolic variation from analysis, increasing the likelihood differences, and obtaining the maximum extraction yield. Low molecular weight metabolites in rat serum were recovered by ELMs using 12 nano-baskets of calixarene, as both emulsifier and carrier. The yields of ELMs were optimized by the method of one-at-a-time. According to NMR data, the maximum metabolic variation was achieved using scaffold 4 (4 wt %), n-decane membrane, stirring rate of 300 rpm, treat and phase ratios of 0.3 and 0.8, respectively. The results revealed that some calixarenes tend to extract non-specific macromolecules; and repeatability of fingerprints for 7-mediated ELM was maximum and for 3-mediated ELM was minimum. The yield of extractions was obtained to be higher for n-decane and lower for carbon tetrachloride. Among different membranes, the fingerprints by chlorinated liquid membranes were more repeatable than using toluene or n-decane.