The relative sensitivities of bioassays to determine the toxicity of sediments were investigated and three methods of making the sample dilutions required to generate dose-response relationships were compared. The assays studied were: (a) Microtox, a 15-min assay of Photobacterium phosphoreum bioluminescence inhibition by pore water; (b) 48-h Daphnia magna lethality test in pore water; (c) 10-d subchronic assay of lethality to and reduction of weight gain by Chironomus tentans performed in either whole sediment or pore water; (d) 168-h acute lethality assay of Hexagenia limbata in either whole sediment or pore water. The three methods of diluting sediments were: (a) extracting pore water from the toxic location and dilution with pore water from the control station; (b) diluting whole sediment from the toxic location with control whole sediment from a reference location, then extracting pore water; and (c) diluting toxic, whole sediment with whole sediment from a reference location, then using the whole sediment in bioassays. Based on lethality, H. limbata was the most sensitive organism to the toxicity of Detroit River sediment. Lethality of D. magna in pore water was similar to that of H. limbata in whole sediment and can be used to predict effects of whole sediment toxicity to H. limbata. (Copyright (c) 1990 SETAC.)