Testosterone (T) and androstenedione (A) metabolism was studied in the polycystic ovary syndrome (PCOS) before and after bilateral wedge resection of the ovaries. In addition, studies of the percentage binding of testosterone in plasma are presented in the PCOS and in normal subjects. The plasma concentrations (PCT), metabolic clearance (MCRT) and plasma production rates (PRT) of testosterone in the PCOS were significantly increased. The PCA and PRA also increased in the PCOS, although the MCRA was within the normal range. The conversion ratios of testosterone and androstenedione and the rate transfer constants of these steroids were not significantly different from the normal group. Plasma androstenedione was the precursor of 14% of plasma testosterone in the PCOS, indicating that the major portion of plasma testosterone arose either from direct glandular secretion or by interconversion from a precursor other than androstenedione. Bilateral ovarian wedge resection resulted in an appreciable fall in the PCT and PRT. This may provide an explanation for the response of patients with the PCOS to this procedure. The percentage binding of testosterone in plasma was significantly lower than that found in normal women and similar to that of normal men. This finding in association with an increased PCT indicates an appreciable increase in the in vivo concentration of unbound steroid in the PCOS. Increasing the concentration of testosterone in normal female plasma in vitro did not result in any appreciable decrease in the percentage binding the hormone until plasma levels of the steroid exceeded the male physiologic range. Since the PCOS is associated with an increase in the PCT to only the lower limits of the male physiologic range, we can conclude that the decreased percentage binding of testosterone in plasma in this disorder is not primarily related to saturation the testosterone binding protein by the increased amount of the circulating steroid. In vitro saturation studies of PCOS plasma demonstrated similarity to male plasma, suggesting that the PCOS may be associated with either qualitative or quantitative changes in the testosterone binding protein or increased amounts of circulating steroids, other than testosterone, which compete for the testosterone binding sites. Increasing the plasma levels of testosterone chronically to the male range in normal women resulted in a dissociation between the percentage binding of testosterone and the MCRT. These findings would be unexpected if protein binding factors were the sole determinant of the MCRT. This suggests that the sex difference in MCRT and the increased MCRT in the PCOS may related, in part at least, to induction of testosterone catabolizing enzymes associated with chronic exposure to increased circulating levels of testosterone.