Genetic backcrosses of C57BL/6 and DBA/2 mice were used to examine the influence of maternal and fetal polymorphisms at the Ahr locus on susceptibility to transplacental carcinogenesis by 3-methylcholanthrene, 7,12-dimethylbenz[a]anthracene, and benzo[a]pyrene, (C57BL/6 x DBA/2) F-1 mothers were backcrossed to DBA/2 males, and DBA/2 females to F-1 males to produce both Ahr-responsive (Ah(+)) and nonresponsive (Ah(-)) fetuses carried by mothers that were themselves either Ah(+) or Ah(-), 3-Methylcholanthrene was given intragastrically on gestation days 13-18 and 7,12-dimethylbenz[a]anthracene or benzo[a]pyrene on day 17 as a single intraperitoneal dose. Ahr phenotype was determined by the zoxazolamine sleeping time test after beta-naphthoflavone pretreatment at 6 weeks of age, The offspring were examined for rumours at 1 year, Both 3-methylcholanthrene and 7,12-dimethylbenz[a]anthracene treatments resulted in a two- to five-fold greater incidence and multiplicity of lung and liver rumours in the Ah(+) offspring compared with that in Ah(-) littermates. By contrast, there was no difference between Ah(+) and Ah(-) offspring with regard to numbers of tumours caused by benzo[a]pyrene, Maternal Ahr phenotype appeared to play a role also, in that the offspring of the Ahr-responsive F-1 mothers developed fewer tumours per unit dose than those of the nonresponsive DBA/2 mothers. The effect of maternal phenotype on risk was three- to five-fold. Fetal and maternal phenotype combined yielded a 10- to 20-fold risk differential for transplacental carcinogenesis by the methylated compounds, with greatest risk experienced by responsive fetuses in nonresponsive mothers, and least by nonresponsive progeny of responsive mothers.