AbstractStudies have been reviewed on the interaction of C14‐ and H3‐labeled carcinogenic hydrocarbons with the tissue constituents of mouse skinin vivo.It has been found that there is a covalent binding between the hydrocarbons and the soluble proteins of mouse skin. Fractionation of these by starch‐gel electrophoresis has shown one characteristic radioactive band of proteins to which the hydrocarbons are bound in a direct quantitative relation to their carcinogenic activity. These proteins resemble electrophoretically the soluble proteins from rat liver to which the hepatocarcinogenie aminoazo dyes are also covalently bound. These proteins are absent from the tumors induced by the hydrocarbons. It has been shown that some papers reporting thein vitrointeraction of carcinogenic hydrocarbons and DNA are in error, and that no binding had actually occurred. In studiesin vivo, no radioactivity could be detected in DNA from mouse skin purified by cesium chloride density gradient centrifugation, following application of highly radioactive hydrocarbons. A possible theory that could explain carcinogenesis by means of a derangement in metabolic control circuitry has been proposed. This theory eliminates the necessity of postulating a direct interaction between the carcinogen and genetic mat