In a previous study, it was demonstrated that mobilization of organochlorine pesticides would be underestimated by an in vitro gastrointestinal model if the sorption of the mobilized pollutants on the digestive residue was not taken into consideration. A multiple fluid/solid ratio procedure was developed to characterize the sorption. In this study, the sorption hypothesis was further tested for polycyclic aromatic hydrocarbons (PAHs), and the sorption of the mobilized PAHs on digestive residue was directly characterized by spiking the gastrointestinal digest with several deuterated PAHs. It was demonstrated that 10-41% of the spiked deuterated PAHs were sorbed on the assimilated residue, which would remain in the solid phase after separation. It appears that the mobility of PAHs would be underestimated if only those dissolved in the fluid is measured. The total mobilized fraction of a PAH compound was defined as a sum of that dissolved in the fluid and that sorbed on the residue. The average mobilized PAH fractions in the studied soils was 70 +/- 24% which was significantly higher than 47 +/- 19% in the fluid. It was also found that the sorption of the mobilized PAHs on the digestive residue was positively correlated with both soil organic carbon (SOC) and molecular weight (MW(t)) of PAHs, and a regression model was developed so that the sorption of different PAHs on soils with different SOCs could be estimated.