Fundamental experiments were conducted with three-phase liquid-phase microextraction (LPME) to optimize greenness, followed by evaluation with the AGREEprep metric system. Extractions were from human plasma, and recoveries were studied for 90 different basic substances (drugs and endogenous metabolites) in the log P-range from -4.2 to 8.1. In a first set of experiments, LPME was performed in a 96-well system using traditional LPME chemicals. Extraction was from 125 µL of human plasma, diluted with 125 µL of 10 mM NaOH, through 3.0 µL of dihexyl ether as liquid membrane, and into 50 µL of 20 mM HCOOH as acceptor. The optimal extraction window (OEW) with this system ranged from log P 2.2 to 5.0, and the majority of drug substances within OEW were extracted exhaustively (average recovery 71%). For this procedure, the AGREEprep metric score for greenness was 0.62. In a next set of experiments, NaOH, dihexyl ether, and HCOOH were replaced with NaHCO3, sesame oil, and citric acid to improve greenness. Even with sesame oil, the majority of the drug substances with log P from 2.2 to 5.0 were extracted with high efficiency, and the AGREEprep score increased to 0.68 due to the highly green nature of the chemicals. In a third set of experiments, where LPME was conducted in a vial-based approach, greenness was improved further by recycling and washing equipment in hot water, except the liquid membrane. Extraction performance was unaffected by washing and recycling, and no carry-over was observed from the extraction equipment. Upon equipment washing and recycling, the total material waste was reduced to less than 10 mg per sample, and the AGREEprep score increased to 0.72. Because experiments were conducted manually and in laboratory, and because LC-MS was used for final analysis, we were unable to improve the AGREEprep score further. The experiments illustrate that LPME procedures may be developed to a very high level of greenness, under assistance by the AGREEprep metric assessment approach.