In this study, we propose procedures based on computational calculations and theoretical models that can be used to predict the behaviour of some of the membrane materials of interest for gas separation applications. In particular, we focus on: i) body-centred cubic VNiTi alloys as novel materials for H2-selective dense membranes and ii) crystalline porous materials that are attractive media for separation of light gases such as H2, O2, CO, CO2, CH4 and N2. These two types of materials are treated using different methodologies, adapted to the needs of our research objectives associated to each material.In the case of dense metal membranes, the long-standing controversy over occupancy of interstitial hydrogen in V-based alloys is addressed. The V-Ni-Ti system is of particular interest here, exhibiting high H2 permeability and improved mechanical properties relative to pure V. This work intends to gain understanding of hydrogen-metal interactions as function of alloy composition and thereby to optimize these new materials and advance their development as novel membranes for H2 separation. We use a first-principles approach that gives insights into the sites preference of hydrogen and assesses the role of Ti and Ni substitutional solutes for the hydrogen absorption affinity. The method based on Density Functional Theory requires no experimental input except crystal structure information. Furthermore, it uses no empirical or fitting parameters in contrast to other computational techniques. Hence this approach provides an alternative way to explore new metal alloys for H2 separation membranes. The applied methodology can be used further in high-throughput calculations to screen various alloy compositions. The hereto-reported results will be used as guidance for tailoring the formulation of VNiTi solid solutions and preparation of low cost dense alloy membranes in the frame of other projects (e.g. European DEMCAMER project).Further, we explore how single-component inputs can be used to forecast the ideal selectivity towards light gases of crystalline porous materials, used for membrane preparation. Theoretical models for describing gas separation properties of zeotype materials as function of structural characteristics and operation conditions are proposed. The model parameters can be obtained as experimentally as well as computationally. To analyse the extent of validity and limitations of the models, ideal selectivities of few crystalline porous materials are evaluated, including widely used zeolites (NaA, CaA) and a metal organic framework structure (ZIF-8). The results verified that the theoretical expressions could be used for screening series of zeotype materials when reliable single gas adsorption data are available. However, since the models don’t take into account all parameters (namely related to the membrane design) and mechanisms involved in gas transport through porous membranes, their predictions should be considered as values referring to an ideal case.