The biological availability of selenium (Se) in barley, oats and meat meal as compared to that in sodium selenite was determined by whole blood glutathione peroxidase (GSH-Px) activity and Se concentration depletion-repletion bioassays. One day old White Leghorn chicks, initially of low-Se status, were further depleted for Se by feeding a low-Se casein basal diet (0.03 mg Se/kg diet) adequate in vitamin E for the first two weeks of life. Repletion treatments of Se at graded levels supplemented by the natural sources or sodium selenite were then provided for 4 weeks. A multiple regression equation was calculated to relate the final whole blood GSH-Px activity and Se concentration to the dietary intake of Se. The percent biological availability of Se present in the natural sources was evaluated by slope ratio analyses in comparison to the standard, sodium selenite (100%). The activity of whole blood GSH-Px established a relative bioavailability for barley, oats and meat meal Se of 85, 41 and 25%, respectively (as mu kat/l), and 81, 51 and 17%, respectively (as mu kat/g Hb). The estimated bioavailability of Se in the different natural sources, as measured by blood Se concentration was: barley, 104%; oats, 99% and meat meal, 69% in comparison to that of sodium selenite. Thus, selenite Se was superior to the natural Se sources for the induction of GSH-Px activity. Barley Se appeared to be of higher biological value to chicks than did oats and meat meal Se. Even though meal Se was retained in appreciable amounts in chick blood, it showed the poorest utilization for the synthesis of blood GSH-Px.