The amphipod Diporeia spp. was exposed to pyrene(0.14 to 1.11 μmol g−1) or phenanthrene (0.08 to 0.62 μmol g−1)-dosed sediments for month-long exposures. Phenanthrene was only slightly toxic with 12±3% mortality at the highest sediment dose (0.62 μmol g−1). Failure to attain and maintain toxic residue body burdens, based on a nonpolar narcosis concentration of approximately 6 μmol g−1, accounts for the low mortality. Phenanthrene toxicokinetic parameters were essentially constant among all doses and consistent with previous measures. Sediment concentration was a poor representation of dose for mortality by pyrene. The relative pyrene distribution among the <63-μm particles increased in the smallest-sized particles at larger doses. An apparent stimulation of pyrene accumulation was observed as a peak in uptake clearance values between sediment concentrations of 0.16 and 0.26 μmol g−1 dry sediment. (Uptake clearance is the amount of source scavenged of contaminant per mass of organism per time.) The pyrene particle-size distribution and the variation in kinetics with dose provide a partial explanation for the poor representation of dose by the sediment concentration. The pyrene body burdens provided a good dose response yielding LD50 values of 6.3 (4.6–41.7, 95% C.I.) and 9.4 (7.9–54.2) μmol g−1 for two experiments. These values are consistent with the residue concentrations for 50% mortality by a nonpolar narcosis mechanism. Comparing the experimental and predicted equilibrium partitioning-based sediment concentrations for 50% mortality, the equilibrium prediction overestimates the toxic pyrene sediment concentration by approximately a factor of ten. Diporeia behavior, differential particle-size distribution, and kinetic limitations appear as likely reasons for the variation between calculated and observed concentrations required to produce mortality.