Recommendations for safe and adequate dietary intakes of selenium were recently established. The recommendations were based largely on data from animals, because few data for humans were available. Some information regarding the dietary selenium intake required by humans to replace excretory losses has now appeared, but the bioavailability to humans of selenium from different dietary sources has not been determined. Selenium bioavailability depends on several metabolic processes, including not only absorption but also conversion into a biochemically active form. Of the different forms of selenium in foods, probably not all are converted to biologically active selenium with equal ease. Therefore, determination of selenium absorption may not in itself yield an accurate estimate of selenium bioavailability. Rather, functional tests, such as measurement of the selenoenzyme glutathione peroxidase, apparently are more valid for determining selenium bioavailability. Various food sources of selenium for humans differ widely in their ability to restore hepatic glutathione peroxidase activity in selenium-depleted rats. Platelet glutathione peroxidase activity seems to be particularly promising for bioavailability studies because platelets are a convenient biopsy material and the platelet enzyme is sensitive to changes in dietary selenium intake. Work in progress should establish the feasibility of this approach for future research in this area.