A brief historical survey is presented on the evaluation of silica adsorbents in analytical HPLC. The theory of analytical HPLC is mostly still being based on the height equivalent to a theoretical plate concept and the van Deemter equation that was derived from gas phase adsorption involving a linear adsorption isotherm and fast mass transfer kinetics. One can obviously wonder whether the use of the van Deemter equation is relevant and valid for the evaluation of the performance of HPLC systems, where most often the liquid solutes involve charged molecules in electrolytes and in very many cases the adsorbates are macromolecules having diffusion coefficients of small magnitude. Instead of the van Deemter equation, a multi-scale modelling approach that involves microscopic and macroscopic dynamic non-linear mass-transfer-rate models should be employed. Furthermore, advanced experimental methods for the characterisation of porous media and the distribution of the density of immobilised active sites (e.g., ligands) on surfaces as well as microscopic pore–network modelling and molecular dynamics modelling and simulation methods could be used for the design of novel adsorbents whose porous structures and immobilised active sites would provide effective mass transport and adsorption rates for realising efficient separations as well as high dynamic capacities when larger throughputs are required.