In search for nonsteroidal inhibitors of 5α-reductase for the treatment of benign prostatic hyperplasia (BPH) and possibly prostate cancer, substrate mimicks were synthesized comprising of a 1-methyl-2-pyridone ( 2, 4–16) or 1-methyl-2-piperidone ( 1, 3, 17–22) moiety (mimicking steroidal ring A) and a diisopropyl ( 1, 2, (E)-5- (E)-7, (Z)-5- (Z)-7, 11–13, 17–19) or a tert-butyl ( 3, 4, (E)-8- (E)-10, (Z)-8- (Z)-10, 14–16, 20–22) benzamide (mimicking steroidal ring D). The bridge connecting the 5 and 4 positions of the rings consisted of amide (CONH: 1–4), ethenyl (CHCH, CCH 3CH, CHCH 3: (E)-5- (E)-10, (Z)-5- (Z)-10) or ethylene groups (CH 2CH 2, CHCH 3CH 2, CH 2CHCH 3: 11–22). The amides 1–4 were obtained by amidation of the carboxylic acid chlorides with the 4-amino- N-substituted benzamides. The ethenyl compounds (E)-5- (E)-10 and (Z)-5- (Z)-10 were synthesized by Wittig reaction of the carbonyl compounds and the corresponding triphenylphosphonium salts and subsequent separation of the stereoisomers. Depending on the time of reaction, catalytic hydrogenation of the ethenyl isomers (E)-5- (E)-10 led to the pyridone-substituted ethylene compounds 11–16 as well as to the piperidone-substituted ethylene compounds 17–22. The 5α-reductase inhibitory properties were determined using rat ventral prostate, as well as human BPH and prostate cancer as source, 1β,2β-[ 3H]testosterone as substrate and a HPLC procedure for the separation of dihydrotestosterone (DHT). Tested at a concentration of 100 μM, the inhibition values of 1–22 ranged from 0–99%. Significant differences were observed between rat and human enzyme. The most active compound was N,N-diisopropyl-4-[2-(1-methyl-2-oxo-piperidine-5-yl)ethylene]benzamide 17 (IC 50: 13 μM).