The P-glycoprotein product (Pgp) of the MDR1 gene has been implicated in the multiple drug resistance phenotype expressed by many cancers. Functioning as an efflux pump, P-glycoprotein prevents the accumulation of high intracellular concentrations of substrates. We have taken a rational approach to designing inhibitors of P-glycoprotein function, selecting a natural substrate (progesterone) as our lead compound. We hypothesized that progesterone, substituted at C-7 with an aromatic moiety(s), would exhibit reduced Pgp affinity, significantly increased antiPgp activity, and reduced affinity for progesterone receptors (PGR). We synthesized 7 alpha-[4'-(aminophenyl)thio]pregna-4-ene-3,20-dione (2), which comprises a C-7 alpha thiol bridge linking an aminophenyl moiety to progesterone, from pregna-4,6-diene-3,20-dione (1). The subsequent addition reaction of 2 with the appropriate isocyanate produced an initial series of compounds (3-6). Compounds 3-5 (respectively, -CH(2)CH(2)Cl; -CH(2)CH(3); and -CH(CH(3))C(6)H(5)) exhibit a significantly increased ability to inhibit P-glycoprotein. Potency for restoring doxorubicin accumulation in MDR1-transduced human breast cancer cells is increased up to 60-fold as compared with progesterone. Compound 5 has greater potency than verapamil and is equipotent with cyclosporin A, for inhibiting P-glycoprotein function. Furthermore, 5 does not bind to PGR, implying a potential reduction in in vivo toxicity. These data identify C-7-substituted progesterone analogues and 5, in particular, as rationally designed antiPgp compounds worthy of further evaluation/development.