Scale-up of aqueous two-phase extraction, which is useful in the isolation and purification of certain bioproducts, is limited by the slow demixing rates of the two aqueous phases. Electrokinetic demixing has been shown to increase by more than 5-fold the demixing rates of systems up to 100 mL in volume in a manner that depends on field strength, field polarity, pH, and phase composition. The present study is an attempt to relate demixing rates to droplet electrokinetic mobilities which were measured microscopically and inferred from demixing data. A clear dependence of demixing rate was observed on drop electrophoretic mobility and pH. The electrophoretic mobility of individual phase droplets suspended in the other phase was measured for poly(ethylene glycol)/Dextran systems using a microelectrophoresis unit and compared with mobilities predicted by electrokinetic theory. We confirmed earlier reports that the droplet electrophoretic mobility increased with increasing drop diameter and explained this increase on the basis of an internal electroosmotic flow model. Effective electrophoretic mobilities were estimated from electrokinetic demixing data in a 100-mL column and compared with predicted as well as experimentally measured values of electrophoretic mobility. The mobilties increased with increased phosphate ionization due to change in pH irrespective of the sign (or polarity) of the applied electric field. The electroosmotic flow model could explain satisfactorily the following two paradoxes: (1) the direction of migration of drops is the opposite of that predicted by colloid electrokinetics and (2) the phase demixing rate increased irrespective of the sign of the applied electric field.