One of the main challenges for the deployment of lignocellulosic biorefineries in future years is to find renewable and secured biomass sources in order to obtain bio-sourced products, as an alternative to petroleum-based commodities. Grass biomass, considering its characteristics (availability, composition, productivity, possibility of being harvested from both arable (post-harvest residues) and non-agricultural lands), can be considered as a biomass source for the future. Nevertheless, because of its complex structure and composition, which need deconstructive pre-treatments to render possible further biological conversions, grasses utilisation in biorefinery is today not widespread. Indeed, recalcitrance to polymers degradation in grasses concerns structural and compositional characteristics and can result in costly and complicated biorefinery processes. Grass recalcitrance is due to various natural factors strongly related and difficult to dissociate: rind and vascular structures; composition (lignin content is a key factor for cellulose hydrolysis acting like a physical barrier while hemicelluloses seem to play a more significant role in woody biomass than in grass plants); physical structures (crystalline nature and insoluble surface of cellulose, specific surface area, particle size), etc. Physico-chemical pretreatments are efficient solutions to overcome recalcitrance, while phenotypic selections are interesting but not efficient enough to obtain an optimal enzymatic hydrolysis. In some cases, the structural elements of grass biomass can be negatively affected by physico-chemical pretreatments, causing pre-treatment-induced recalcitrance, like cellulose hornification (irreversible alteration of cellulose microfibers), vascular structure collapsed and reduced cellulose bioaccesibility to enzymes due to cellulose covering by lignin, following lignin solubilisation.