To elucidate the relationship between binding parameters and biopotencies of glucocorticoids, we partially purified the receptor from the liver cytosol of rats in a dexamethasone-bound and unactivated form by precipitation with protamine sulfate, gel filtration and DEAE-cellulose chromatography (approximately 100-fold) and examined the interaction of the preparation with 3 glucocorticoids of different biopotencies (dexamethasone; Dex, corticosterone; Cort and prednisolone; Pred). The partially purified receptor (PPR) was stable at -20 degrees C for at least 2 months in the presence of bovine serum albumin, glycerol, molybdate and dithiothreitol. Treatment of the PPR with p-hydroxymercuribenzoate liberated the ligands and the treated PPR reassociated 3H-glucocorticoids efficiently following the addition of dithiothreitol. The reassociated PPR was bound to the DNA-cellulose after a brief heating. Metabolic activity on ligands and inactivation of the binding sites in the PPR were insignificant under the conditions used. Kd's were approximately 0.9, approximately 3 and approximately 6 nM for Dex, Cort and Pred, respectively (at 0 degree C). Relative binding affinity of ligands to the PPR which was estimated by competitions was higher in the order of triamcinolone acetonide greater than Dex greater than Cort greater than Pred greater than progesterone greater than cortexolone. Association of Dex and Cort was relatively rapid and significantly accelerated by raising the incubation temperature, while the association of Pred was slower and effects of the temperature was moderate. The rate of dissociations was also varied with ligands. The rate of dissociation of Dex was the lowest among the 3 ligands and was elevated by raising the temperature. Because the effect of temperature was more pronounced in the dissociation than in the association, apparent Ka's decreased at higher temperature. Thermodynamic examinations of glucocorticoid binding in the PPR revealed that the binding reaction proceeds at a higher rate in the order of Dex greater than Cort greater than Pred. Because the relative biopotencies of these 3 glucocorticoids in vivo is higher in the order of Dex greater than Pred greater than Cort, from the results obtained in the present study, it appears that biopotency of glucocorticoids in vivo does not correlate with the affinity of the binding to the receptor estimated in vitro.