Insulin-like growth factor (IGF)-binding protein-3 (IGFBP-3) is the principal carrier protein of IGF-I in the circulation. IGF-I has been postulated to play a role in the genesis and maintenance of the polycystic ovary syndrome. Regardless of the exact mechanism of action of IGF-I on ovarian steroidogenesis, alterations in the level of IGFBP-3 may play a significant role in regulating the concentration of IGF-I in hyperandrogenism. We have postulated that androgens reduce the circulating IGFBP-3 concentration through a mechanism similar to its suppression of the hepatic production of sex hormone-binding globulin (SHBG), thereby increasing bioavailable IGF-I and amplifying its impact on ovarian steroidogenesis. To test this hypothesis, we studied seven oophorectomized women (aged 39-51 yr; body mass, 20.9-35.8 kg/m2) during 3 weeks of testosterone (T) propionate administration (20 mg, three times weekly). All subjects were receiving 0.625 or 1.25 mg conjugated estrogens/day. Blood was sampled before (week 0), during (weeks 1-3), and after (week 4) T administration. Serum was assayed for total T, GH, and SHBG, and plasma was assessed for IGF-I, insulin (INS), and IGFBP-3. IGFBP-3 was measured by both RIA and Western ligand blotting; (expressed as a percentage of the control value). Circulating T increased from 1.51 ± 1.06 to 30.8 ± 13.8 mmol/L by week 2 (P < 0.001). During T administration, IGF-I increased (from 55 ± 23 ng/mL at week 0 to 124 ± 37 ng/mL at week 4; P < 0.05); INS did not change, with the exception of a higher fasting level 1 week after discontinuing therapy, and GH decreased (from 1.7 ± 2.3 µgrams/L at week 0 to 0.4 ± 0.4 µgram/L at week 4; P < 0.03), as did the circulating SHBG concentration (397 ± 205 to 273 ± 93 nmol/L by week 2; P < 0.01). IGFBP-3 levels determined by Western ligand blot were higher during the second and third weeks of T administration (265 ± 28% and 218 ± 43% of control values, respectively; P < 0.05) compared to that at week 0 (165 ± 44% of control values). However, there was no difference in the circulating concentration of IGFBP-3, determined by RIA, at weeks 0, 1, 3, and 4 (3.59 ± 0.35, 4.00 ± 0.79, 3.48 ± 0.56, and 3.65 ± 0.52 µgram/mL, respectively). 23.4 rg/mL at week 0 to 123.9 f 37.3 pg/mL at week 4; P < 0.05), and the immunoreactive IGF-I to IGFBP-3 ratio increased linearly (from 16.1 -C 5.7 at week 0 to 27.6 + 8.7 at week 4; P < 0.03). In conclusion, an increase in the circulating T concentration did not suppress circulating IGFBP-3 levels, as occurred with SHBG. Nonetheless, a linear increase in the IGF-I to IGFBP-3 ratio was noted, suggesting that free IGF-I may increase with androgen excess. The clinical significance of these tindings in the maintenance of the polycystic ovary syndrome remains to be determined, because the T levels achieved in this study were well above those normally found in affected women. Finally, the impact of androgens on hepatic production/secretion appears to differ for SHBG vs. IGFBP-3. (J Clin Endocrinol Metub 79: 1747-1751,1994). © 1994 by The Endocrine Society.