The work presented in this thesis has focused on the development and implementation of two-dimensional (2D) nuclear magnetic resonance (NMR) correlation techniques to unambiguously discriminate and characterise competitive adsorption and mass transfer processes in porous materials. This has primarily involved investigations on porous oxides; in particular silicas and aluminas commonly used as catalyst supports. The techniques used are demonstrated to be capable of acquiring information relevant to the performance of heterogeneous catalysts and adsorbents across the hierarchy of length scales relevant to industrial processes. The methodologies associated with the acquisition and processing of 2D NMR correlation data were first established through the development of analytical models capable of simulating 2D signal attenuation data for T1-T2, D-T2 and T2-T2 experiments. Common artefacts were also discussed by means of experimental and simulated examples and, where appropriate, methods have been introduced for their prevention. The NMR relaxation behaviour of water saturating the pore space of silica was observed to correlate strongly with independent measurements of the activation energy of dehydroxylation, thus establishing NMR relaxometry as a tool for directly probing the surface energetics of silica surfaces. This interpretation of T1/T2 ratios differs from that in conventional applications of the technique which typically present the ratio as an indicator of surface-adsorbate interaction strength. Here, the T1/T2 ratios of three liquid probe molecules: ethanol, diethyl ether and cyclohexane, are used to investigate the influence of pore size and density of adsorption sites on relaxation behaviour. Competitive adsorption behaviour has been directly investigated through the acquisition of T1-T2 correlation data of binary liquid mixtures imbibed in various silica supports. These measurements, in combination with a newly developed model for the relaxation of multi-component mixtures, have provided a comprehensive assessment of the ability of this technique to quantify intra-pellet compositions and address competitive adsorption behaviour in porous media. With the aid of a random walk Monte-Carlo model to simulate transverse relaxation in packed bed reactors, T2-T2 relaxation exchange measurements have been used to investigate mass transfer across the fluid-solid boundary in a packed bed reactor filled with γ-alumina and liquid cyclohexane. These data were used to quantify the rate of exchange between intra- and inter-pellet environments at a number of flow rates. Exchange rates were then converted into the more convenient terms of mass transfer coefficients and compared against literature data using two separate dimensionless mass transfer analyses.