The bioavailability of polycyclic aromatic hydrocarbons (PAHs) for earthworms (Eisenia andrei) was experimentally determined in seven field-polluted soils and 15 soil-sediment mixtures. The pore-water concentration of most PAHs was higher than predicted. However, most of the compound was associated with dissolved organic carbon (DOC) and not directly available for uptake by earthworms. The apparent sorption could be reasonably predicted on the basis of interactions with DOC; however, the biota-soil accumulation factors (BSAFs) for earthworms were up to two orders of magnitude lower than predicted by equilibrium partitioning. The large variability between sites was not fully explained by differences in sorption. Experimental results indicate that the pool of freely dissolved PAHs in the pore water became partially depleted because of uptake by the earthworms and that bioaccumulation is thus also influenced by the kinetics of PAH desorption and mass transport. A pilot study with Lumbricus rubellus showed that steady-state body residues were well correlated to E. andrei. Current results show that depositing dredge spoil on land may lead to increased bioavailability of the lower-molecular-weight PAHs. However, risk assessment can conservatively rely on equilibrium partitioning, but accurate prediction requires quantification of the kinetics of bioavailability.