Background: Improving lignocellulolytic enzymes' diffusion and accessibility to their substrate in the plant cell walls is recognised as a critical issue for optimising saccharification. Although many chemical features are considered as detrimental to saccharification, enzymes' dynamics within the cell walls remains poorly explored and understood. To address this issue, poplar fragments were submitted to hot water and ionic liquid pretreatments selected for their contrasted effects on both the structure and composition of lignocellulose. In addition to chemical composition and porosity analyses, the diffusion of polyethylene glycol probes of different sizes was measured at three different time points during the saccharification. Results: Probes' diffusion was mainly affected by probes size and pretreatments but only slightly by saccharification time. This means that, despite the removal of polysaccharides during saccharification, diffusion of probes was not improved since they became hindered by changes in lignin conformation, whose relative amount increased over time. Porosity measurements showed that probes' diffusion was highly correlated with the amount of pores having a diameter at least five times the size of the probes. Testing the relationship with saccharification demonstrated that accessibility of 1.3-1.7-nm radius probes measured by FRAP on non-hydrolysed samples was highly correlated with poplar digestibility together with the measurement of initial porosity on the range 5-20 nm. Conclusion: Mobility measurements performed before hydrolysis can serve to explain and even predict saccharification with accuracy. The discrepancy observed between probes' size and pores' diameters to explain accessibility is likely due to biomass features such as lignin content and composition that prevent probes' diffusion through nonspecific interactions probably leading to pores' entanglements.