Serious problems of interpretation may arise when metabolic studies are carried out with radioactively labeled drugs and other foreign chemicals. Detection of free or bound radioactivity in tissues or body fluids may indicate the presence of the unchanged chemical or of various products of its decomposition. Some radioactivity may, however, represent incorporation of certain metabolites of foreign chemicals into body constituents by normal biosynthetic pathways. The interpretation of the results of such metabolic experiments from the standpoint of safety evaluation will be profoundly different if the radioactivity represents covalent binding to a cellular macromolecule than if it results from normal endogenous incorporation. These two distinct types of binding of radioactivity to body constituents are well illustrated by experimental studies with some carcinogenic nitroso compounds, such as dimethylnitrosamine and N-methylnitrosourea. Dimethylnitrosamine requires metabolic activation by microsomal enzymes and yields formaldehyde and a chemically reactive methylating intermediate, probably a methyl carbonium ion. The latter reacts with nucleophilic sites in nucleic acids and proteins, and also with water to yield methanol. N-Methylnitrosourea does not require metabolic activation but yields the same methylating intermediate spontaneously under physiologic conditions. Both formaldehyde and methanol are metabolized largely to CO2, but they also enter the one-carbon metabolic pool and became biosynthetically incorporated into nucleic acids, proteins, and other cell components. Alkylation of cellular constituents is associated with various biological effects, including cytotoxicity, carcinogensis, and mutagenesis, and the same effects are produced by the activated forms of a variety of other chemical carcinogens. It is clearly of paramount importance to distinguish between these two types of incorporation of radioactivity.