The transport dynamics for two membrane based separation processes, forward osmosis (FO) and facilitated diffusion (FD), are studied. A three-pronged approach is pursued to assist in the design of these two systems: an analytical model describes performance as a function of membrane properties and operating parameters; numerical solutions of the governing equations corroborate the model; and bench-scale experimental modules demonstrate the ability of the model to predict performance. The effective operation of FO and FD modules requires that operating conditions be chosen based on the membrane and treated streams; the derived analytical models are vital in aiding these design choices.FO processes are an emerging set of technologies that show promise in the treatment of complex and impaired water streams. In this study, the derived forward osmosis model ascertains membrane selectivity and draw solute concentration as the dominant factors that regulate the extent of water recovery and solute rejection. This model expresses the performance (e.g., recovery rate and separation factors) in terms of operating conditions (e.g., the flow rates and concentrations of the inlet solutions) and membrane transport properties (e.g., the hydraulic and solute permeability coefficients).FD processes could enhance the extraction of deleterious dilute contaminants from water supplies; in these processes, membranes separate molecules based on chemical factors, rather than physical factors. In this study, polymeric membranes that exhibit FD are investigated. An analytical model that relates the performance (e.g., selective solute transport) to membrane properties (e.g., thickness, binding capacity) and operating parameters (e.g., feed solute concentration, oscillation frequency) of these inherently unsteady-state systems, is developed. The solute binding capacity and thickness of the membrane define the asymptotic limit for enhanced selectivity in these systems. Therefore, thin membranes that possess high solute binding capacities are needed to push the development of these chemically-active membranes further. The analytical models derived for FO and FD processes are reinforced by numerical algorithms that solve their respective governing equations and by experimental results that demonstrate the models ability to predict performance. Specifically, multifunctional membranes for FD processes (which consist of a responsive gate layer that is coated onto a reactive matrix) and cellulose acetate membranes for FO modules (which are commercially available) are used to corroborate the assertions derived from their respective model.