A method for the assessment of both selenium-dependent and non-selenium-dependent glutathione peroxidases on crude tissue extracts in human is described. The enzyme activity is measured by the coupled assay system in which oxidation of reduced glutathione (GSH) is coupled to NADPH oxidation catalyzed by glutathione reductase. Total glutathione peroxidase activity is measured with cumene hydroperoxide as substrate. Selenium-dependent glutathione peroxidase is measured with tert-butyl hydroperoxide. This substrate is preferable to H2O2 which gives too high blank values compared to the assay values. The difference between total glutathione peroxidase and selenium-dependent glutathione peroxidase activities represents the non-selenium-dependent glutathione peroxidase activity. Studies of substrate specificity of the two glutathione peroxidases separated by gel filtration as well as linearity and recovery studies are presented. For a given tissue, the relative amounts of the two glutathione peroxidases given by our assay are identical to those estimated by quantifying the elution peaks after gel filtration. Based on the percentages of the two glutathione peroxidases, human tissues can be classified in four groups: (1) the non-selenium-dependent glutathione peroxidase is predominant in liver, in renal cortex and skeletal muscle; (2) non-selenium-dependent and selenium-dependent glutathione peroxidases are in equal amounts in renal medulla; (3) the selenium-dependent glutathione peroxidase is predominant in adrenal glands and platelets; (4) the selenium-dependent glutathione peroxidase represents 100% of the glutathione peroxidase activity in the other organs. The heart and the brain are of special interest in this group because of the physiological role and the regulation of the selenoenzyme.