Homogenates of liver, lung, kidney, stomach, small intestine and colon from 8 strains of mice were compared for their ability to metabolize benzo[a]pyrene (BP) and dimethylnitrosamine (DMN) to mutagens. Females of strains CF1, AKR/J, AU/SsJ, DBA/2J, SWR/J, A/J, C3H/HeJ, and C57BL/6J were either untreated or received phenobarbital (PB), 3-methylcholanthrene (MC) or polychlorinated biphenyls (AR) to induce drug-metabolizing enzymes. The effects of these drugs on organ weight and on the amounts of DNA, S-10 protein, and microsomal protein per unit weight of tissue are reported. Salmonella typhimurium TA92 and TA98 were used as indicators of the formation of mutagens. For each organ there was an optimal balance between amount of tissue homogenate and concentration of test compound for maximal yield of revertants. A sensitive radiometric assay of DMN demethylase (DMND) is described which permits measurement of the enzyme in liver, lung and kidney. DMN at 1 mM is used as substrate. Aryl hydrocarbon hydroxylase (AHH) was measured in all tissue using BP as substrate. AR and MC are very good inducers of AHH activity in livers of mice classified as aromatic hydrocarbon responsive, but not in those classified as hydrocarbon nonresponsive. Responsiveness is strain-specific and genetically regulated. Metabolism of BP to mutagens by liver homogenates was correlated with extent of AHH induction. This dimorphism of response of AHH to inducers was present, but less pronounced, in non-hepatic tissues. Basal activities of AHH and DMND were correlated in livers and lungs from untreated mice. DMND activities were increased less than 2-fold by PB, MC or AR treatments. Metabolism of DMN to mutagens was not closely correlated with DMND activities. Strain of mouse, type of tissue and test substance are important variables in assessing the potential effect of microsomal enzyme-inducing agents on the metabolism of mutagenic substances.