RECENT results by Strong1,2 and by Carr3 on mice, and by Demerec4 on Drosophila, have focused interest on a possible connexion between mutagenicity and carcinogenicity, which in the past has often been claimed on theoretical grounds. While Auerbach5 failed to produce mutations by injection of 1 : 2 : 5 : 6 dibenzanthracene into Drosophila, Demerec obtained positive results with a number of carcinogens, including dibenzanthracene and methylcholanthrene, when these were administered to the flies in the form of aerosols. Still a different possibility of administration is by addition of the carcinogen to the food of the larvae, since fluorescence tests carried out by Fabian and Matoltsy6 have shown that methylcholanthrene given in this way enters into the gonads. An experiment using this method was under way when Demerec‘s first results Were published, and its outcome will be briefly presented. The food contained powdered methylcholanthrene in a concentration of 0·01 per cent. No live yeast was added to the culture vials, in order to avoid complications through offering the larvae an alternative supply of untreated food. Fluorescence tests showed that methylcholanthrene or a fluorescent derivative of it had penetrated into the gonads of the imagines which emerged in the treated vials. In spite of this, no evidence for mutagenic action of methylcholanthrene was obtained in tests for sex-linked lethals. Only two lethals were found in 1,012 chromosomes from treated ♂♂ (compared with one in 1,007 control chromosomes), and three in 1,026 chromosomes from treated ♀♀ (compared with two in 1,014 control chromosomes).