The chemist is much intrigued with attempts to discover relationships, in test animals, between carcinogenic activity and molecular structure of benz[a]anthracene and related compounds. A Unified hypothesis that all chemical preprocarcinogens of aromatic structural type ArX (X = H) undergo some biochemical substitution reaction to yield chemical procarcinogens (X = H) and proximate carcinogens (X = OH) with aromatic structural type ArCH₂X was tested. In perfect agreement with this hypothesis, all non-methylated meso-region positions of preprocarcinogens are methylated by S-adenosyl-L-methionine-dependent methyltransferases. The meso-methyl group of the procarcinogen is hydroxylated, in vivo, to form the proximate carcinogen, which, in turn, is converted to the ultimate carcinogen by conjugation with a good leaving group. The accumulated evidence shows that carcinogenic meso-aldehydes and ketones are reduced by aldo-keto reductases to benzylic alcohols. The ultimate carcinogens, having structural type ArCH₂X such as hydroxymethyl sulfate esters derived by PAPS-dependent sulfotransferases and benzylic halides, are powerful aralkylating agents capable of generating benzylic carbocations, benzylic DNA adducts, and cancer. This, then, is the alternative to current views that radical cations or diols, derived from epoxides, as precursors of diol-epoxide or quinone metabolites, together with reactive oxygen species, are not carcinogenically inert. It has been suggested that the oxidized metabolites of structural type ArX (X = one or more oxygen atoms) are not mutually exclusive, and that the presence in vivo of all exceptionally reactive metabolites of type ArX causing DNA damage may be required to account for the pronounced carcinogenic properties of chemical preprocarcinogens and procarcinogens.