Steady-state performances of a trickle bed reactor (TBR) and a biofilter (BF) in loading experiments with increasing inlet concentrations of polar solvents, acetone, methyl ethyl ketone, methyl isobutyl ketone and n-butyl acetate, were investigated, along with the system's dynamic responses. Throughout the entire experimentation time, a constant loading rate of aromatic components of 4 g(c)·m(-3)·h(-1) was maintained to observe the interactions between the polar substrates and aromatic hydrocarbons. Under low combined substrate loadings, the BF outperformed TBR not only in the removal of aromatic hydrocarbons but also in the removal of polar substrates. However, increasing the loading rate of polar components above the threshold value of 31-36 g(c)·m(-3)·h(-1) resulted in a steep and significant drop in the removal efficiencies of both polar (except for butyl acetate) and hydrophobic components, which was more pronounced in the BF; so the relative TBR/BF efficiency became reversed under such overloading conditions. A step-drop of the overall OL(POLAR) (combined loading by polar air pollutants) from overloading values to 7 g(c)·m(-3)·h(-1) resulted in an increase of all pollutant removal efficiencies, although in TBR the recovery was preceded by lag periods lasting between 5 min (methyl ethyl ketone) to 3.7 h (acetone). The occurrence of lag periods in the TBR recovery was, in part, due to the saturation of mineral medium with water-soluble polar solvents, particularly, acetone. The observed bioreactor behavior was consistent with the biological steps being rate-limiting.