Selenium, an essential trace element, has been shown to decrease plasma glucose concentrations of diabetic rats. To study the short-term effects of selenium on hepatic carbohydrate metabolism, isolated perfused livers of fed Sprague-Dawley rats were continuously infused with sodium selenite for 90 minutes. This resulted in an immediate elevation of selenium in the effluent perfusate (3.3 +/- 0.1, 16.1 +/- 0.4, 30.3 +/- 1.6, and 118.9 +/- 0.8 mumol/L at infusion of 10, 50, 100, and 500 mumol/L sodium selenite, respectively). Basal hepatic glucose production decreased in a dose-dependent manner within 60 minutes of low-dose sodium selenite infusion (10: 0.60 +/- 0.20, 50: 0.21 +/- 0.40, and 100 mumol/L: 0.21 +/- 0.09 mumol.min-1.g-1 liver; P < .05 vs. zero time), while it was transiently increased by 500 mumol/L sodium selenite (1.11 +/- 0.18 mumol.min-1.g-1 liver; P < .05). Glucagon-stimulated glycogenolysis was suppressed by 50% (P < .05) at 1.8 nmol/L insulin and by 90% (P < .001) at 10 mumol/L sodium selenite. That selenium concentration did not affect glutathione peroxidase activities in liver and perfusate erythrocytes within 60 minutes. Toxic effects of high-dose selenite (500 mumol/L), but not of low-dose selenite (10 mumol/L) infusion, were indicated by increased hepatic glucose (P < .05), lactate (P < .01), and lactate dehydrogenase (P < .001) release as well as histologically by degeneration and necrosis of periportal hepatocytes. In conclusion, low-dose selenite exerts a potent insulinlike effect on hepatic glycogenolysis in vitro by counteracting glucagon action, whereas high-dose selenite may severely impair liver function.