It is known that 19-nor-deoxycorticosterone (19-nor-DOC) is a potent mineralocorticosteroid that is present in urine of rats and humans in a free, i.e., nonconjugated, form. In some forms of hypertension in rats, the levels of free 19-nor-DOC in urine are increased compared with those in urine of normotensive animals. Yet, despite the potential importance of this mineralocorticosteroid in the pathogenesis of certain forms of hypertension, little is known of its site of origin or metabolism. In the present investigation, we evaluated the metabolism of intravenously infused [3H]19-nor-DOC and the possibility that 19-nor-DOC was formed from plasma DOC. We found that the metabolism of [3H]19-nor-DOC infused intravenously in men and women was similar to that of DOC with important exceptions. The majority of the radiolabeled urinary metabolites of intravenously infused [3H]19-nor-DOC were excreted in urine as glucuronosides. Little radioactivity, infused as [3H]19-nor-DOC, was recovered in urine as nonconjugated or sulfoconjugated steroids. There was no free radiolabeled 19-nor-DOC in urine after the simultaneous infusion of [3H]19-nor-DOC and [14C]DOC. A major metabolite of [3H]19-nor-DOC in urine was 19-nor-DOC-21-glucuronoside, whereas little or no intravenously infused radiolabeled DOC was excreted as radiolabeled DOC-glucuronoside. We also found that intravenously infused [14C]DOC was not converted to urinary [14C]19-nor-DOC (glucuronoside) and that other tritium-labeled metabolites of infused [3H]19-nor-DOC contained no carbon-14. The production rate of 19-nor-DOC, computed from the specific activity of urinary 19-nor-DOC (glucuronoside), in one normal man was 16 micrograms/d and in the two women of this study, it was 10 micrograms/d. These findings are supportive of the proposition that free urinary 19-nor-DOC is not formed from plasma DOC; it may be formed in kidney from a precursor other than DOC or it may be formed nonenzymatically in kidney or urine from a precursor such as 19-oic-DOC.