1. 1. The carcinogenic aromatic amine 2-aminofluorene (AF), labeled with 14C in carbon atom 9, became readily bound to the proteins of aerobically incubated rat liver slices. The isotope became most rapidly bound to the microsomal proteins. The binding was inhibited to some extent when the slices were incubated with amino acid analogues such as ethionine and p-fluorophenylalanine. At the same time the activity of oxidative demethylase, one of the enzymes for oxidative detoxication in the microsomes, decreased in the slices. 2. 2. Under similar conditions isotope from 14C-aniline became bound to the proteins of the slices. This simple, non-carcinogenic amine was bound, however, to a considerably smaller extent than 14C-AF. 3. 3. In the presence of a TPNH-generating system isotope from 14C-AF became bound to the proteins of a rat liver homogenate. The microsomal components were of essential importance for this binding. In the presence of TPNH the isotope became bound also to the proteins of isolated microsomes. TPNH could to a certain, limited extent be replaced by DPNH. The binding process showed a pH-optimum at 8–8.5. The binding reaction was not significantly inhibited by the amino acid analogues, but quite strongly by sulfhydryl reagents such as dihydroxydinaphthyldisulfide. It was also inhibited, probably in a competitive way, by glutathione and cysteine. Since, moreover, a considerable part of the bound radioactivity was released by treatment of the incubated proteins with thioglycollic acid, it is considered that sulfhydryl groups were involved to a considerable extent in this binding process. 4. 4. A binding of isotope from 14C-aniline to the proteins of a rat liver homogenate was observed under similar conditions. The aniline-binding was constantly lower than the binding of 14C-AF, and the relative difference increased by the addition of a sulfation system to the incubation mixture. 5. 5. It is believed that reactive oxidation intermediates are formed in microsomal structures, when AF is metabolized. The occurrence of such metabolites may involve certain consequences, which are discussed particularly with respect to the protein metabolism. The possibility is considered that reactive metabolites of liver carcinogens may act as cytoplasmic mutagens.