Human pubic skin was obtained from normal subjects and patients with abnormal sex differentiation. Skin samples (200 mg) supplemented with NADPH, were incubated for 1 h with labelled testosterone. The conversion of testosterone to dihydrotestosterone1), 3\g=a\-and 3\g=b\-androstanediol was calculated. This conversion averaged 14.9 \m=+-\ 3.4 % (se) in 11 normal men and 3.6 \ m=+-\1.4 % (se) in 8 normal women. In 4 children as in 4 young hypogonadotrophic hypogonadal men, the conversion rate of testosterone to 5\g=a\-reducedmetabolites was low (0.8 to 3.5%) and increased at puberty (13.5 to 19.2%). After administration of HCG for 3 months to 1 of the hypogonadal men, it reached 30.2 %. Inversely, the formation of dihydrotestosterone and androstanediols from testosterone was suppressed in 2 men treated with large doses of oestrogen. In 3 subjects with an incomplete form of testicular feminization syndrome, the conversion rate of testosterone to 5\g=a\-reducedmetabolites was in the normal male range (6.4 to 18.3%), whereas it was low in one case of the complete form of the syndrome (1.5%). In 9 women with idiopathic hirsutism the rate of 5\g=a\-reducedmetabolites recovered from testosterone was close to that of normal men (13.5 \ m=+-\5.5% (se). From theseresults, it is postulated that in human subjects, there is a good correlation between hair growth in skin from a sexual area and the extent of testosterone 5\g=a\-reductionin this tissue. Such an enzymatic activity 0 The following trivial names have been used: Dihydrotestosterone = 17/?-hydroxy-5aandrostan-3-one; androstenedione = 4-androstene-3,17-dione; 3ci-androstanediol = 5aandrostane-3«,17/?-diol; 3/î-androstanediol = 5«-androstane-3/5\17/?-diol; androstanediols = 3a-androstanediol + 3/i-androstanediol.