Twenty mature geldings, averaging 535 kg, were used to determine the influence of dietary selenium (Se) on the blood levels of Se and Se-dependent glutathione peroxidase (SeGSH-Px) activity in the horse. Horses were randomly assigned within breed to four treatments consisting of five horses each and fed a basal diet containing .06 ppm of naturally occurring Se. Diets were supplemented with .05, .10 and .20 ppm Se, as sodium selenite. Blood was drawn for 2 wk before, and for 12 wk following, the inclusion of supplement Se in the diets. Whole blood and plasma Se concentrations and plasma SeGSH-Px activities were determined from all blood samples. Selenium concentrations in plasma and whole blood increased linearly from wk 1 to wk 5 and 6, respectively, in Se-supplemented horses. After these times, no significant changes in Se concentration were observed in Se-supplemented or in unsupplemented horses throughout the remainder of the 12-wk trial. Plasma Se reached plateaus of .10 to .11, .12 to .14, and .13 to .14 micrograms/ml in horses supplemented with .05, .10 and .20 ppm Se, respectively. Whole blood Se reached plateaus of .16 to .18, .19 to .21, and .17 to .18 micrograms/ml in horses supplemented with .05, .10 and .20 ppm Se, respectively. Plasma SeGSH-Px activity was not significantly affected by dietary treatment. Therefore, this enzyme was not a good indicator of dietary Se in these mature horses.