Organically modified clay minerals have been widely developed, tested and employed as sorbents for organic pollutants. However, the process of pollutant-composite pairing is not commonly addressed, which would be valuable for efficient pollutant filtration by such sorbents. This study presents an approach for achieving efficient pollutant removal by large-scale composite filters, based on pairing chemically compatible pollutants and composites and by employing a predictive filtration model. The removal of three organic pollutants, simazine, sulfentrazone and diclofenac by lab-scale filtration columns containing one of three sorbents, a polymer-, micelle- or liposome-clay composite, was measured. Understanding the factors governing pollutant-organic modifier interactions enabled to pair an efficient sorbent to each pollutant. The high removal (80%) of simazine by the polymer composite, was attributed to hydrogen bonds and π-π interactions, compared to less than 20% removal by the surfactant composites. The removal of the anionic diclofenac (pKa=4.1) was mainly governed by electrostatic attraction, explaining its high removal by the most positively charge sorbent, the liposome composite. Sulfentrazone (pKa=6.5) removal was mostly affected by micellar solubilization and upon its removal, the zeta potential of the micelle-composite was not reduced as obtained for diclofenac removal. The filtration of the successful pairs was modelled to determine sorbent capacity and adsorption and desorption rate constants. The pilot filtration experiments were well described by the model and demonstrated efficient removal of paired pollutants and sorbents. Model simulations predicted promising treatment at environmental pollutant concentrations in the μgL−1 range. This pairing approach along with model calculations can be a strong and valid tool for efficient pollutant-sorbent filtration.