The hexacyclic aromatic hydrocarbon dibenzo[def,p]chrysene, better known as dibenzo[a,l]pyrene (DBP) in the field of chemical carcinogenesis, is present in the environment as a combustion product of organic matter. This compound is probably the strongest chemical carcinogen ever tested. As ultimate genotoxic metabolites of DBP two electrophilically reactive species are discussed: (i) radical cations generated by one-electron oxidation, and (ii) fjord region dihydrodiol epoxides formed via the trans-11,12-dihydroxy 11,12-dihydro derivative of DBP (11,12-dihydrodiol). In order to delineate the metabolic pathway(s) involved in tumor formation by DBP, newborn Crl:CD-1(ICR)BR mice were intraperitoneally treated with the parent compound, its 11,12-dihydrodiol, and the two diastereomeric fjord region dihydrodiol epoxides. Due to severe acute and chronic toxicity, the total dose of DBP and of the 11,12-dihydrodiol was limited to 40 nmol. For the same reason the dihydrodiol epoxides could only be applied in doses up to 0.4 nmol. The tumor incidence was determined 55 +/- 1 weeks after treatment. Under these conditions, DBP and its 11,12-dihydrodiol induced lung tumors (incidence: 86.5% versus 92.0%; yield: 2.88 versus 7.44 tumors per mouse), liver (incidence: 57.7% versus 60.0%; yield: 3.63 versus 5.28 tumors per mouse) and other organs (incidence: 36.5% versus 32.0%; yield: 0.56 versus 0.52 tumors per mouse). By contrast, only lung tumors at low incidence were detected in mice treated with solvent only (incidence: 28.8%; yield: 0.58 tumors per mouse). As with the parent hydrocarbon, mice treated with low doses of diastereomeric syn- and anti-dihydrodiol epoxides of DBP showed increased tumor incidences in liver (incidence: 19.0 and 46.7%; yield: 0.36 and 1.47 tumors per mouse, respectively), and in various other organs (incidence: 7.1 and 20.0%; yield: 0.07 and 0.20 tumors per mouse, respectively). In consideration of the 100-fold differences in the doses of compounds applied in this study, the tumor-inducing potency increases in the order DBP < 11,12-dihydrodiol < anti-dihydrodiol epoxide. This result provides strong evidence that the potent carcinogen DBP is activated in vivo in the mouse via its 11,12-dihydrodiol and not preferentially through alternative pathways.