Lignocellulosic biomass represents a valuable candidate for the replacement of fossil resources in the production of value-added chemicals, materials and biofuels, thanks to a diverse composition rich in carbohydrates (i.e. cellulose and hemicellulose) and aromatics (i.e. lignin). However, the success of lignocellulose as the feedstock of tomorrow relies on the valorisation of its full content, which requires appropriate methodologies for fractionation into its components and, particularly, the preservation of lignin reactivity. The lignin-first biorefining based on Early-stage Catalytic Conversion of Lignin (ECCL) is among the most promising approaches for catalytic fractionation of biomass. It deconstructs biomass in organic solvents in the presence of a hydrogenation catalyst, enabling the recovery of a high-quality carbohydrate fraction together with a highly functionalised lignin product. This dissertation investigates the mechanistic aspects underpinning the Catalytic Upstream Biorefining process (CUB) based on the ECCL concept, employing Raney Ni as the hydrogenation catalyst and 2-PrOH 7:3 v/v as a lignin-extracting liquor and H-donor. Finally, the biomass structure played a crucial role in the transport of the released lignin fragments from the lignocellulosic matrix to the reaction medium, where catalytic reactions can occur, as higher deconstruction achieved in CUB experiments performed in more severe conditions enhanced the degrees of lignin depolymerisation. In the context of biomass deconstruction, sugars derived from hemicellulose degradation in CUB were found to have a relevant impact on Raney Ni's catalytic selectivity, which could be shifted from phenols to cyclohexanols by enhancing the preservation of hemicellulose. An investigation on the time evolution of the physicochemical processes occurring in CUB highlighted the relevance of the first 50 min of the process, during which the extent of delignification is the highest. Importantly, the activity of Raney Ni in limiting repolymerisation reactions on the released lignin fragments is maintained throughout the time range 0-300min. Through a Design of Experiment study, the interplay of process parameters (i.e. wood loading, catalyst loading, solvent volume) on the performances of CUB (i.e. pulp delignification, hemicellulose retention in pulp, ratio low-MW/high-MW species in lignin, total content of monophenols in lignin, content of hydroxypropyl phenols in lignin) was assessed in detail, revealing that the amount of catalyst employed and the concentration of lignin species in the reaction medium relevantly affect biomass deconstruction and lignin depolymerisation. Based on the acquired knowledge, optimum conditions for maximising the performances of CUB were designed and tested, showing promising results for the application of the process to a large scale.