Three tests are routinely done to assess blood status of selenium in cattle: serum selenium, whole blood selenium, and glutathione peroxidase. The objective of this study was to compare the various analytical methods for determining blood selenium status in groups of mature cows and beef calves. Twenty to 30 blood samples per herd were collected from 8 beef herds in central Alberta and 1 dairy in Alberta herd twice a year from the spring of 1992 through the fall of 1995, and once from 185 spring calves in 2 beef herds in Saskatchewan. Serum and whole blood samples were submitted to 1 laboratory and whole blood samples were submitted to a 2nd laboratory. Samples for glutathione peroxidase determinations were submitted to a 3rd laboratory. Pearson's correlation coefficients and Cohen's kappa were calculated for each possible comparison among the different measures. The best agreement was observed between serum and whole blood analysis within Laboratory A. The remaining comparisons reflected poor agreement. Comparison of herd-level assessment resulted in better agreement than comparison of individual sample results among laboratories and procedures for all combinations tested. Serum selenium analysis was the only laboratory procedure for which external reference material was utilized. Serum selenium, whole blood selenium, and glutathione peroxidase measure different compartments of the blood selenium pool. The time frame of interest, supplementation practices, and the stability of recent dietary intake determine the optimum assessment method for individual animals or herds. Determination of the serum status or of blood selenium is more consistently measured at the herd-level than for individual samples.