A new family of non-steroidal 5-alpha-reductase inhibitors was designed by replacing the steroid skeleton of an inhibitor related to estrone by a biphenyl moiety. This hypothesis originated from the reported estrogenic activity of a few biphenyl compounds (see Part 1 of this paper; Lesuisse et al. Bioorg. Med. Chem. Lett. 2001, 11, 1709). Two compounds turned out to be potent type 2 5-alpha-reductase inhibitors with IC(50)'s of inhibition in the nanomolar range. These are to our knowledge amongst the most potent non-steroidal 5-alpha-reductase inhibitors described to date.