A number of N-nitroso compounds and an azoxyalkane have been labeled with deuterium in various positions and have been administered to rats, hamsters, or mice in parallel with the unlabeled compounds. The treatments with the labeled and analogous unlabeled compounds were equimolar and for the same time. Mortality rates from tumors and tumor incidences were compared between deuterium-labeled and the unlabeled analogs. In many cases more than one dose level was used for the comparisons. An increased rate of mortality from tumors or an increased incidence of induced tumors was considered an index of increased potency of one treatment compared with the other. Using these criteria deuterium in the alpha positions of nitrosodimethylamine, nitrosomorpholine, nitrosoheptamethyleneimine, and nitrosoazetidine reduced carcinogenic potency compared with the unlabeled compounds. This indicated that cleavage of a carbon-hydrogen bond in the alpha position was a rate-limiting step in carcinogenesis by these nitrosamines. In both nitrosomethylethylamine and nitroso-2,6-dimethylmorpholine, the presence of deuterium at different positions increased or decreased carcinogenic potency, suggesting that competition for oxidation between these sites might be the determining factor in activation of the molecule. This also applied to nitrosomethyl-n-butylamine and nitrosomethyl-phenylethylamine with deuterium at the methyl group or at the alpha carbon of the butyl or phenylethyl groups, and to azoxymethane with deuterium in the 1-methyl or 4-methyl group. In nitrosomethylcyclohexylamine, nitrosomethyl-n-dodecylamine, and dinitroso-2,6-dimethylpiperazine there was no detectable effect of deuterium on carcinogenic potency, suggesting that the conditions did not provide sufficient sensitivity for detection of an isotope effect, or that oxidation at the alpha carbon was not a rate-limiting step in carcinogenesis by these molecules.