The desorption and bioavailability of 3,3′,4,4′-tetrachlorobiphenyl (PCB 77) were studied in spiked natural sediments at six concentrations. The desorption kinetics were measured in a sediment–water suspension using Tenax® resin extraction, and the bioavailability was measured by exposing Lumbriculus variegatus (Oligochaeta) to PCB 77–spiked sediment in a 14-d kinetic study. In addition, freely dissolved pore-water concentrations were measured using the polyoxymethylene solid-phase extraction method. The present study examined whether bioavailability can be defined more accurately by measuring the size of desorbing fractions and the pore-water concentrations than by using the standard equilibrium partitioning approach. The importance of ingested sediment in bioaccumulation also was investigated. Our data showed a clear, decreasing trend in the rapid-desorbing fractions and in the standard biota–sediment accumulation factors (BSAF) with increasing concentration in sediment. Desorbing fractions–refined BSAFs were more uniform across the concentration treatments, and the pore-water PCB 77 concentration predicted tissue concentrations close to observed values. In the risk assessment process, pore-water concentration or desorbing fractions would lead to more precise bioavailability estimates compared with those from the traditional equilibrium partitioning approach. The result also showed, however, that sediment-ingesting worms had access to an additional bioavailable chemical fraction that was especially evident when PCB 77 pore-water concentrations most likely approached the solubility limit. Thus, feeding may modify the bioavailable fraction that cannot be explained by simple equilibrium partitioning models.