Selenium is present in plasma and tissues in specific and non-specific forms. The experiments reported here were carried out to clarify some factors that affect these forms of the element in plasma. A selenium-replete human subject was given 400 microg of selenium daily for 28 days as selenomethionine and, in a separate experiment, as selenate. The selenomethionine raised plasma and albumin selenium concentrations. Selenate did neither. The molar ratio of methionine to selenium in albumin was approximately 8000 under basal and selenate-supplemented conditions but 2800 after selenomethionine supplementation. This demonstrates that selenium from selenomethionine, but not selenium from selenate, can be incorporated into albumin, presumably as selenomethionine in the methionine pool. Selenocysteine incorporation into albumin was studied in rats using (75)Se-selenocysteine. No evidence was obtained for incorporation of (75)Se into albumin after exogenous administration or endogenous synthesis of (75)Se-selenocysteine. Thus, selenocysteine does not appear to be incorporated non-specifically into proteins as is selenomethionine. These findings are in support of selenomethionine being a non-specific form of selenium that is metabolized as a constituent of the methionine pool and is unaffected by specific selenium metabolic processes. No evidence was found for non-specific incorporation of selenium into plasma proteins when it was administered as selenate or as selenocysteine. These forms of the element appear to be metabolized by specific selenium metabolic processes.