Elevated concentrations of planar, halogenated hydrocarbons have been linked to reproductive problems in a variety of fish-eating birds and mammals in the Great Lakes and in particular Saginaw Bay. Currently, there are no accurate procedures to assess bioavailability of these contaminants. Polychlorinated dibenzo-p-dioxins and dibenzofurans and mono- and non-ortho-chloro-substituted biphenyls in water at the femtogram to picogram per liter range were passively concentrated in semipermeable membrane devices (SPMDs), and these data were compared to the bioconcentration in co-exposed (caged) channel catfish. Sediment-derived water concentration estimates, calculated from a steady-state partitioning model, did not correlate well to those derived from either fish or SPMDs. The use of SPMDs demonstrated the utility of in-situ passive sampling over inference of water concentrations from accumulation in biota or partitioning with sediment. Residues ac cumulated by SPMDs have been shown to be proportional to analyte water concentration, whereas this does not appear to be the case for fish tissues. The greater amounts of 3,3‘,4,4‘-tetrachlorobiphenyl and 2,3,7,8-tetrachlorodibenzofuran accumulated in SPMDs than in exposed channel catfish indicated those non-passive aspects of bioconcentration in organisms, such as biotransformation and elimination, introduced 50−500% error in the assumed degree of exposure.