This work investigated selective micro-electromembrane extractions (μ-EMEs) of the colored indicators metanil yellow and congo red (visual proof-of-principle) and the small drug substances nortriptyline, papaverine, mianserin, and citalopram (model analytes) based on their acid-base strength. With two free liquid membranes (FLMs), the target analytes were extracted from aqueous donor solution, across FLM 1 (1-pentanol, 1-ethyl-2-nitrobenzene (ENB) or 4-nitrocumene (4-NC)), into aqueous acceptor solution 1, further across FLM 2 (1-pentanol, ENB or 4-NC), and finally into aqueous acceptor solution 2. All phases had volumes between 1.0 and 1.5 μL and extractions were promoted by 200-300 V d.c. applied across the five-phase μ-EME system formed in a perfluoroalkoxy capillary tubing. The anode was located in acceptor solution 2 and the cathode was located in donor solution for μ-EMEs of acidic analytes, and locations of the electrodes were vice versa for μ-EMEs of basic analytes. After μ-EME, donor solution and acceptor solution 1 and 2 were analyzed by capillary electrophoresis or liquid chromatography-mass spectrometry. The model analytes migrated efficiently in the proposed μ-EME system, their migration behavior was controlled by pH in aqueous solutions and their selective fractionation into acceptor solution 1 and 2 was demonstrated based on their acid-base strength. Under optimal conditions, acceptor solution 2 contained 60% nortriptyline (pKa = 10.5) and less than 1% papaverine (pKa = 6.0) and acceptor solution 1 contained 17% nortriptyline and 27% papaverine after 15 min of μ-EME. The five-phase μ-EME system was also compatible with human plasma samples. Work is in progress to further increase the fractionation capability, and to implement the concept into microfluidic platforms.