Selenium has been reported to affect glutathione (GSH) concentrations in short-term animal-feeding experiments. Given the central role that this tripeptide plays in maintaining cellular homeostasis, it was hypothesized that perturbations in glutathione metabolism induced by selenium might account for its cancer chemopreventive activity. In the present study, four experiments were conducted in which the effect of acute, short-, or long-term exposure to selenium was assessed. Selenium was provided as either sodium selenite or D,L-selenomethionine. Selenite was observed to induce a biphasic response in total liver GSH. Injected selenium caused an acute reduction in GSH, whereas short-term feeding (up to 8 wk) increased both total GSH and oxidized glutathione (GSSH), an effect that gradually diminished in magnitude with prolonged feeding. Our data suggest that such changes are unlikely to account for the chemopreventive activity of selenium for the following reasons: Perturbations in glutathione metabolism occurred only at doses of selenite that approached toxicity. These doses are higher than what would be required for producing cancer chemoprevention. The transient nature of these changes also contrasts with the need for a continuous supplementation of selenite in suppression of tumorigenesis. Furthermore, selenomethionine was found to have little activity in altering glutathione metabolism, even though it compares favorably with selenite as a cancer chemopreventive agent. Nonetheless, these findings do not discount the possibility that sulfhydryl compounds, such as glutathione, might be used to modify the toxicity and/or enhance the cancer prophylactic activity of selenium compounds.