Dibenz[a,j]acridine (DBAJAC) was studied because of its close structural relationship with a number of important carcinogenic polycyclic and azaaromatic hydrocarbons. It was of particular relevance to examine the mutagenicity of known or proposed 'bay-region' metabolites, which may be proximate or ultimate carcinogenic derivatives of DBAJAC. Trans-1,2-, 3,4- and 5,6-dihydrodiols, the 4- and 6-phenols, the 5,6-oxide and N-oxide derivatives, and anti- and syn-3,4-diol 1,2-epoxides of DBAJAC were examined for their mutagenicity in Salmonella typhimurium TA98 and TA100 and in V79 Chinese hamster lung cells. Of all the compounds studied which require metabolic activation, the 3,4-dihydrodiol was the most active in both TA100 and in V79 cells. The activity of the 3,4-dihydrodiol enantiomers was also tested in strain TA100 where no difference was observed from that of the racemic mixture. In V79 cells only the 3R,4R-dihydrodiol was active, the activity being about three times that of the racemic material. Salmonella strains TA98 and TA100 also differed in their sensitivity towards DBAJAC dihydrodiols, the 1,2-isomer being of greatest activity in TA98. The most mutagenic compounds in both mammalian and bacterial cells were the 'bay-region' diol epoxides of DBAJAC which did not require metabolic activation by S9 mix. The anti-DBAJAC 3,4-diol 1,2-epoxide was more mutagenic than the syn form in V79, TA98 and TA100 cells. Overall these results suggest that the in vivo biological activity of DBAJAC metabolites is likely to reflect previous findings with other similar polycyclic aromatic hydrocarbons.