Hydrogen separation membranes could be an enabling technology in a hydrogen economy. A comprehensive microkinetic model for hydrogen permeation was developed to expand the ability to predict hydrogen flux through various potential dense metal (solid and liquid) membrane candidates over a wide range of operating conditions. The molecular steps in the assumed mechanism, i.e., surface adsorption, dissociation, infiltration, and bulk diffusion, are adopted from the literature. The limiting assumptions made normally in the literature models, however, were avoided to develop a more comprehensive and rigorous model while still being computationally accessible. The use of an electric circuit analogy to model the molecular permeation step network allowed individual steps to be analyzed insightfully and in identifying which are rate-limiting steps under various conditions and which may be considered to be at quasi-equilibrium. The model was validated using experimental flux data available in the literature, and involving kinetic and thermodynamic parameters derived from theoretical and experimental sources, for the conventional solid palladium and palladium- silver membranes. In order to extend the model to evaluate the efficacy of a sandwiched liquid metal membrane (SLiMM), a novel membrane under development in this laboratory, the molecular step kinetic and thermodynamic parameters must first be determined. Experimental as well as theoretical work was thus performed to determine these parameters for liquid gallium (Ga) and liquid indium (In), two potential SLiMM candidates. A semi-theoretical approach termed the Pauling Bond Valence-Modified Morse Potential (PBV-MMP) method was used to determine activation energies as well as pre-exponential factors for the steps involved in hydrogen permeation in a liquid metal. Experimentally, absorption isotherms as well as adsorption and diffusion kinetics were measured using a Sieverts apparatus that operates by measuring the pressure difference when a valve is opened between an evacuated sample chamber of known volume and another chamber charged to an initial pressure of hydrogen gas. Based on the hence theoretically and experimentally determined parameters, the microkinetic model was extended to SLiMM and conditions identified when different steps are rate-limiting or at quasi-equilibrium. The model was compared to experimental data for permeation of hydrogen in liquid Ga and liquid In membranes, giving a reasonable first prediction of the hydrogen flux through each metal membrane, and confirming their potential as hydrogen membranes.