Glucocorticoids have three distinct therapeutically relevant effects (genomic, specific nongenomic, and unspecific non-genomic), raising the hypothesis that the relative potencies of non-genomic and genomic effects of glucocorticoids may differ. Therefore, we measured the unspecific non-genomic potencies of five clinically important glucocorticoids and compared them with the classical (genomic) potencies. We studied the immediate glucocorticoid effects on respiration, on protein synthesis, and on Na+-K+-ATPase and Ca2+-ATPase in concanavalin A-stimulated rat thymocytes. We titrated the respiration of the cells with methylprednisolone, prednylidene, dexamethasone, prednisolone or betamethasone, and then interpolated the glucocorticoid concentrations needed to inhibit concanavalin A-stimulated respiration back to normal. These "equivalent doses" produced equal inhibition of respiration, of specific energy-consuming pathways, and of the concanavalin A effect on quiescent cells. The relative drug potencies were calculated as the inverse of the equivalent doses normalized to methylprednisolone and were: prednylidene (3.0) > dexamethasone (1.2) > methylprednisolone (1.0) > prednisolone (0.4) > betamethasone (0.2). This hierarchy is completely different from that for the classical effects. These new data are of crucial relevance for in vitro experiments and clinical use, especially in glucocorticoid high-dose therapy. Examples are the choice between methylprednisolone and prednisolone in pulse therapy, and the completely different clinical usage of dexamethasone and betamethasone, despite their similar affinities for nuclear receptors.