A chromatographic method is described to determine the distribution of selenium between selenoprotein P, glutathione peroxidase (GSH-Px), and albumin in plasma, using two small columns of heparin-Sepharose and reactive blue 2-Sepharose linked together in tandem. One milliliter of plasma was diluted to 12 ml with 0.02 M sodium phosphate buffer, pH 7.0 (the equilibration buffer), applied to the heparin-Sepharose column, and eluted at a flow rate of 30 ml per hour. GSH-Px was not retained by either of these columns but selenoprotein P was retained by heparin-Sepharose and albumin by reactive blue. After the two columns were separated, selenoprotein P was eluted with heparin from heparin-Sepharose and albumin eluted from reactive blue with high salt. Analytical work confirmed the presence of selenoprotein P, GSH-Px, and albumin in the respective fractions. When rats were injected with 75Se as either selenite or selenomethionine most of the radioactivity was incorporated into the selenoprotein P fraction, with the next greatest amount into GSH-Px, and the least amount into albumin. Slab gel electrophoresis was used to determine that most of the selenium in each of the three fractions was associated with each of these selenium containing proteins. This method indicated that the majority of the selenium in plasma is associated with selenoprotein P, and the only time this was found not to be true was with high levels of dietary selenomethionine.