The present study characterized the DNA adducts induced by manufactured gas plant residue (MGP) and benzo[a]pyrene (B[a]P) in mouse lung and forestomach. The dose levels used in the present study were comparable to the levels used in a previous animal bioassay. Adduct formation was evaluated in female A/J mice (7 weeks old) fed MGP (0.25%) or B[a]P (16 and 98 ppm) for 14 days. In addition, adduct formation was also evaluated in mice 24 h after the ip administration of 1.8 mg of B[a]P in 0.5 mL of tricaprylin. P-32-Postlabeling combined with multidimensional TLC and reverse phase HPLC was used to evaluate hydrocarbon-DNA adducts. HPLC separation of chemical-DNA adducts formed in lung following MGP ingestion resulted in three distinct peaks of radioactivity eluting at 22, 32.4, and 33.5 min. These peaks accounted for 13, 10, and 41% of the total adducts detected. The adducts isolated from forestomach eluted as a series of minor peaks with two more distinct peaks of radioactivity at 32.4 and 33.5 min. These peaks accounted for 47 and 32% of the total adducts detected in forestomach, respectively. Ingestion of B[a]P (16 or 98 ppm) and the ip administration of B[a]P resulted in a single major adduct with a retention time of 32.4 min. The DNA adducts formed from MGP administration were further characterized by comparison with adducts formed following the administration of individual hydrocarbons and a mixture of hydrocarbons. Comparison of adduct retention times formed by MGP and pure hydrocarbons indicates that two of the three adducts formed in lung (22 and 32.4 min peaks) are derived from benzo[b]fluoranthene and B[a]P, respectively. However, the major adduct in lung could not be attributed to any of the PAH identified as constituents of MGP. In contrast, B[a]P is responsible for the major adduct formed in forestomach. These results indicate that several PAHs may play a role in the mechanism by which MGP induces tumors in mouse lung.