The selenoenzyme glutathione peroxidase cannot account for all the physiological effects of selenium in rat liver. Therefore, a study was carried out with the ultimate aim of identifying selenoproteins other than glutathione peroxidase. The incorporation of 75Se, given as 75SeO32-, into centrifugally separated fractions of selenium-deficient and control rat livers was determined. In selenium-deficient liver much less 75Se was incorporated into the 105,000g supernatant fraction than in controls, so this fraction was studied further by gel filtration, ion-exchange, and hydroxylapatite chromatography. Selenoglutathione peroxidase and another selenoprotein, called 75Se-P, were separated and identified. Both these selenoproteins were also found in plasma. Selenium deficiency had opposite effects on incorporation of 75Se by these proteins. It decreased 75Se incorporation by glutathione peroxidase at 3 and 72 h after 75Se injection but increased 75Se incorporation by 75Se-P. This suggests that 75Se-P competes for available selenium better than does glutathione peroxidase when the element is in short supply. Apparent molecular weights of 75Se-P from liver and plasma determined by gel filtration were, respectively, 83,000 and 79,000, which indicate proteins smaller than glutathione peroxidase. Cycloheximide pretreatment of the rat blocked 75Se incorporation into plasma 75Se-P. These experiments establish the existence of a selenoprotein, 75Se-P, in rat liver and plasma which is chromatographically distinct from glutathione peroxidase and which incorporates 75Se differently from glutathione peroxidase. 75Se-P may account for some of the physiological effects of selenium.