Production of volatile mutagenic metabolites from 5 halogenated promutagens was examined by a simple modification of the conventional Salmonella/microsome mutagenicity assay. This method incorporates the taping together of 2 agar plates face to face during the initial portion of their incubation at 37C. By varying the contents of the soft agar in each of the two plates with respect to promutagen. S9 and tester strain cells, mutagenesis due to volatile promutagens and their metabolites could be quantitated separately. Using the taped plate assay, volatile mutagenic metabolites were detected from the promutagens 3-(2-chloroethoxy)-1,2-dichloropropene, the herbicides diallate, triallate and sulfallate, and the flame-retardant tris-(2.3-dibromopropyl) phosphate (Tris-BP). All compounds except Tris-BP were also found to be volatile promutagens. The mutagenic metabolites accounted for 50-80% of the activity of these compounds observed in the standard assay. Moreover, our studies suggest that a small, but appreciable percentage of the mutagenic metabolites from all 5 compounds escaped detection in the conventional, untaped assay. Mutagenic activity of the volatile mutagenic metabolites from diallate was quenched by various Salmonella tester strains independent of their responsiveness to diallate mutagenesis. Detection of volatile mutagen formation from diallate was also prevented by cysteine and glutathionine, but not by DNA or metyrapone. This taped plate method for the Salmonella assay should facilitate future investigations of the detection, isolation and identification of volatile mutagenic metabolites from other promutagenic compounds or mixtures.