Human placental microsomes were incubated with [3H]benzo[a]pyrene (BP) and Salmon sperm DNA and the resulting metabolite-nucleoside complexes resolved by Sephadex LH-20 chromatography. The metabolite pattern was analyzed by high-pressure liquid chromatography (HPLC). The incubates were also co-chromatographed with extracts obtained from incubates with rat liver microsomes and [14C] BP. Phenols, quinones and 7,8-dihydrodiol were detected in the placental incubates. Both 9,10- and 4,5-dihydrodiols were very low as compared with control rat liver samples. Placental microsomes catalyzed the binding of BP metabolites to DNA in vitro, giving rise to two main complexes which co-chromatographed with rat liver-produced peaks attributable to 7,8-diol-9,10-epoxide and 7,8-oxide and/or quinones when metabolized further. The nucleoside metabolite peaks attributable to 4,5-oxide and 9-phenol-4,5-oxide were lacking when compared with the binding pattern catalyzed by rat liver. Both the total binding and specific metabolite-nucleoside adducts in the placenta correlated with fluorometrically measured aryl hydrocarbon hydroxylase (AHH) activity and with the amount of dihydrodiol formed. The results demonstrate that both the metabolite pattern and the nucleoside-metabolite complexes formed by the placental microsomes in vitro differed greatly from thos produced by rat liver microsomes. These sstudies also suggest that it is not possible to predict specific patterns of DNA binding from AHH measurements or even from BP metabolite patterns, especially when comparing different tissues and species.