It has been suggested that ligand-dependent gene activation by the progesterone receptor (PR) can result from recruitment of PR by the promoter bound Sp1. A detailed investigation of the Sp1-dependent agonistic activity of RU486 and R5020 on the folate receptor (FR) type alpha, p27, thymidine kinase 1 and p21 genes reveals a different mechanism. The FR-alpha P4 promoter and the endogenous FR-alpha gene were up-regulated by the PR agonist R5020 through either PR-A or PR-B. The classical antagonist RU486 also activated the promoter but only through PR-B. The most proximal (essential) G/C-rich (Sp1 binding) element and the initiator region constituted the minimal promoter responsive to PR regulation; substitution with a stronger cluster of G/C-rich elements enhanced the magnitude of the PR response. In contrast, substitution of the G/C-rich element with a TATA box resulted in the loss of regulation by PR. Overexpression of Sp1 and Sp4 but not Sp3 enhanced activation of the FR-alpha promoter by PR, knocking down Sp1 decreased the activation in a manner that was reversed by ectopic Sp1 or Sp4. The ligand-dependent action of PR on the promoter was delayed compared with its activation of a classical glucocorticoid response element-driven promoter and activation of both the promoter and the endogenous FR-alpha gene by PR required new protein synthesis. Activation by PR paralleled RNA polymerase II recruitment but was not accompanied by either association of PR or a change in the association of Sp1 with the endogenous FR-alpha P4 promoter. Similar observations were made for PR regulation of the genes encoding p27, thymidine kinase 1, and p21. The results contradict the current view of Sp1-dependent gene regulation by PR and point to the existence of one or more PR target genes whose promoter and cell context(s) must thus be key determinants of the agonistic activity of RU486 on a large group of important Sp1-dependent downstream target genes.