Formaldehyde was examined for bacterial mutagenicity using Escherichia coli WP2(pKM101) and WP2uvrA(pKM101), and Salmonella typhimurium TA1535, TA1537, TA1538, TA98, TA100 and TA102, in the absence of any exogenous source of metabolic activation. Using pre-incubation exposure, clear mutagenicity was seen for TA98, TA100 and TA102, and both E. coli strains. In standard plate-incorporation assays, consistent mutagenicity was seen only for TA100 and WP2uvrA(pKM101). No evidence of mutagenicity was seen for TA1535, TA1537 or TA1538 using either method of exposure. These data confirm the enhanced ability of the pre-incubation method to detect the mutagenicity of formaldehyde both quantitatively, as expressed by numbers of revertant colonies, and qualitatively, in terms of the range of indicator strains reverted. The relatively greater sensitivity of the pre-incubation assay is probably due to better containment of a volatile agent and/or lack of interaction with agar during the initial period of exposure. The findings are consistent with the suggestion that formaldehyde induces lesions in bacteria which are, at least to some extent, excision-repairable, and indicate that the presence of the R-factor plasmid may be required for the expression of its mutagenicity in excision repair-deficient Salmonella.