The effects of selenium on 7,12-dimethylbenzanthracene-induced mammary tumorigenesis were examined in C57BL X DBA/2f F1 mice fed a semipurified diet. Mice fed 0.2 ppm selenium developed 56% mammary tumors; in contrast, mice fed 2.0 ppm selenium developed only 16% mammary tumors at 11 months of age. Mice fed the 2.0-ppm selenium diet grew as well as did their counterparts fed the 0.2-ppm selenium diet. In a separate experiment, the level of selenium-dependent glutathione peroxidase was measured in the mammary glands of control and 7,12-dimethylbenzanthracene-treated BALB/c mice fed basal and selenium-supplemented diets. 7,12-Dimethylbenzanthracene treatment resulted in decreased glutathione peroxidase activity n mice fed both low (0.03 ppm)- and high (1.50 ppm)-selenium diets. Thus, the chemopreventive effects of selenium could not be attributed to maintaining high levels of glutathione peroxidase. In a second series of experiments, the effects of selenium were further examined on the growth of mammary cell line YN-4 in monolayer cell culture. The mitochondrial inclusions seen in cells exposed to 5 X 10(-6) M selenium could not be correlated with changes in the activity of the mitochondrial enzymes, cytochrome c oxidase and succinate dehydrogenase, thus implying that there was no demonstratable impairment of mitochondria. The examination of selenium-treated cells with flow cytofluorometry indicated that cells were blocked in S-G2 phases of the cell cycle. This latter result illustrates one feasible approach towards identifying specific mechanisms for the chemopreventive effects of selenium.