We compared the renal metabolism of Ba-fluorinated steroids with that of the unfluorinated, endogenous steroid cortisol (F). By defining kinetic variables, we characterized isoenzyme activities of 11 beta B-hydroxysteroid-dehydrogenase (11 beta-HSD). Methods: I) In human kidney slices, we studied the conversion of 9 alpha-fluoro-cortisol (FF) and F to their ore-products (and vice versa). II) In human kidney microsomes, we performed the kinetic analysis of 11 beta-HSD activity for the steroid pairs F/cortisone (E) and dexamethasone (D)/11-dehydrodexamethasone (DH-D). Results: I) In kidney slices, FF is very weakly oxidized to 9 alpha-fluorocortisone (FE), while the reduction of FE to FF is very effective. In contrast, E is hardly reduced to F, but F is strongly inactivated to E. II) Enzyme kinetics in kidney microsomes: 1a) Oxidation of F to E: exclusively NAD-dependent; Km = 25.5 nmol/L. b) Reduction of E to F: clearly NADH-preferring; Km = 81 nmol/L; Vmax(oxidation)/Vmax(reduction) (F/E)= 26. 2a) Oxidation of D to DH-D: exclusively NAD-dependent; Km = 81 nmol/L b) Reduction of DH-D to D: exclusively NADH-dependent; Km = 68 nmol/L; Vmax(oxidation) /Vmax(reduction) (D/DH-D) = 0.09. Thus, the equilibrium of FF/FE in human kidney slices is far on the biologically active hydroxy-side. This shift, induced by the 9-fluorination, gives a good explanation for its mineralocorticoid potency. The cosubstrate dependence and the Km-value of the oxidation of F are similar to those of the cloned human 11 beta-HSD-II. For the first time, we could show a NADH-dependent reduction of E to F. Moreover, we found that the preference of D/DH-D for the reductase reaction (see the quotients Vmax(oxidation)/Vmax(reduction)) is due to a NADH-dependent enzyme (probably 11 beta-HSD-II). These results provide strong evidence against the ''dogma'' of an ''unidirectional'' 11 beta-HSD-II.