A factorial experimental design was used within the study to evaluate the influence of multiple metabolic activation system concentrations on the dose-response exhibited by promutagens (indirect-acting mutagens) in the Salmonella spiral and plate assays. The mutagenic activity of the three compounds used spanned three orders of magnitude. The mutagenic activity of the compounds ranged from 10 to 100 revertants/microgram for acetylaminofluorene (2AAF) to more than 1000 revertants/microgram for 2-aminoanthracene (2AA). Benzo(a)pyrene (BaP) activity was within an intermediate range (100-1000 revertants/microgram). During a single experiment, a mutagen was tested in TA100 at 13 doses plus a negative control dose. Each dose was tested at 10 S9 concentrations. The S9 concentrations ranged from 0.1 mg protein/plate to 4 mg protein/plate in the standard plate assay and from 0.25 to 4.90 mg-equivalents in the spiral assay. The spiral Salmonella assay, an automated version of the standard assay, generates dose-response data from a concentration gradient on a single agar plate, thereby providing a straightforward approach to this type of study. The study demonstrates not only that even small differences in S9 concentrations can affect the measurement of mutagenic potency but that S9/compound interactions cannot be generalized through the use of interaction studies. (Copyright (c) 1991 Elsevier Science Publishers B.V.)