Cytosol prepared from cultured AtT-20 mouse pituitary cells or mouse liver was treated with concentrations of p-chloromercuriphenyl sulfonate (PCMPS) which reduced but did not abolish receptor-binding activity. Scatchard analysis of triamcinolone acetonide binding to the treated cytosol showed that the PCMPS effect was caused by a reduction of binding affinity with little effect on the apparent binding site concentration. The effect on affinity was dose-dependent. Binding specificity appeared unaffected since the relative abilities of triamcinolone acetonide, dexamethasone, cortisol, progesterone, and corticosterone to compete with labeled triamcinolone were similar at various PCMPS concentrations which caused a progressive reduction of detectable cytosol binding. The PCMPS effect was reversible since cytosol treated with up to 200 microM PCMPS followed by dithiothreitol 15 min later showed nearly complete recovery of binding sites (62-100%). The possibility that several sulfhydryl groups were involved in this phenomenon was further explored in experiments using AtT-20 cytosol labeled with [3H]dexamethasone-mesylate, a glucocorticoid affinity label which binds covalently to sulfhydryl groups. Chromatography of dexamethasone-mesylate labeled receptor on a sulfhydryl affinity column resulted in binding, indicating that the receptor had at least two sulfhydryl groups, one bound to the mesylate moiety of the steroid and the other capable of binding to the affinity column.