Structured catalysts, such as solid foams, represent a very promising technology for continuous and stable production of high-value compounds derived from biomass, traditionally produced with batch and semibatch technologies using suspended catalysts. However, the synthesis of structured catalysts presents additional challenges related to their structure and the generation of porous coatings with suitable properties for dispersing the catalytically active phase on the support. This work was focused on synthesizing a Ru/C solid foam catalyst and investigating its activity in the selective hydrogenation of xylose to xylitol under different operational conditions. The carbon coating, the key step of preparation, was based on the formation and pyrolysis of poly(furfuryl alcohol) in the presence of different amounts of poly(ethylene glycol) (PEG; M = 8 kDa) as a pore former, which enabled tuning the support porosity. Thus, the catalyst prepared with 5 wt% PEG presented a micro-to-mesopores volume ratio of 1, and a good dispersion of Ru nanoparticles, as well as a better stability compared to the catalyst prepared without PEG. The extensive kinetic data collected in this work were mathematically modelled using three different approaches to elucidate the reactant adsorption mode: a non-competitive adsorption model, a non-competitive adsorption model considering the effect of temperature, and a semi-competitive adsorption model. The non-competitive temperature-dependent model displayed better performance in terms of fitting and reliability of the estimated parameters and predicted the adsorption of xylose as an endothermic process. On the other hand, the semi-competitive model gave similar results in terms of fitting and a value for the competitiveness factor of 0.74, which matches the hypothesis that the larger molecules, sugars, can occupy most of the active sites, while some interstitial sites remain accessible for hydrogen adsorption. The modelling results revealed a complex mode of sugar adsorption on the catalyst surface. This modelling concept can be applied to any system in which the molecule sizes are very different.