These studies were undertaken to investigate (a) the permeability properties of the blood-brain barrier (BBB) to the major gonadal and adrenal steroid hormones, and (b) the role of the binding proteins of plasma (albumin and specific globulins) in the regulation of BBB steroid hormone transport. The permeability of the BBB to [(3)H]-labeled progesterone, testosterone, estradiol, corticosterone, aldosterone, and cortisol, was measured relative to [(14)C]butanol, a freely diffusable reference, in the barbiturate anesthetized rat using a tissue sampling-single injection technique. The isotopes were rapidly injected in a 200-mul bolus of Ringer's solution (0.1 g/dl albumin) via the common carotid artery and the percent extraction of unidirectional influx of hormone was determined after a single pass through brain: progesterone, 83+/-4%; testosterone, 85+/-1%; estradiol, 83+/-3%; corticosterone, 39+/-2%; aldosterone, 3.5+/-0.8%; and cortisol, 1.4+/-0.3%. The selective permeability of the BBB was inversely related to the number of hydrogen bonds each steroid formed in aqueous solution and directly related to the respective 1-octanol/Ringer's partition coefficient. When the bolus injection was 67% human serum, >95% of the labeled steroid was bound as determined by equilibrium dialysis. However, the influx of the steroids through the BBB was inhibited by human serum to a much less extent than would be expected if only the free (dialyzable) hormone was transported; progesterone, estradiol, testosterone, and corticosterone transport was inhibited 18, 47, 70, and 85% respectively, or in proportion to the steroid binding to plasma globulins. Rat serum (67%) only inhibited the transport of these four hormones, 0, 13, 12, and 69%, respectively, reflecting the absence of a sex hormone-binding globulin in rat plasma. However, neonatal rat serum (67%) inhibited progesterone, testosterone, and estradiol transport 0, 0, and 91%, respectively, consistent with the presence of an estradiol-binding protein in neonatal rat serum. The binding of steroid hormone to bovine albumin in vitro (as determined by equilibrium dialysis) was compared to albumin binding in vivo (as determined by the single injection technique). The ratio of apparent dissociation constant in vivo, K(D)(app), to the in vitro K(D) was: >>200 for progesterone, >200 for testosterone, 120 for estradiol, and 7.7 for corticosterone. Assuming the steady-state condition, the K(D)(app)/K(D) was found to be proportional to the BBB permeability for each steroid. These data demonstrate (a) the selective permeability properties of the BBB to the major steroid hormones is proportional to the tendency of the steroid to partition in a polar lipid phase and is inversely related to the number of hydrogen bond-forming functional groups on the steroid nucleus; (b) the presence of albumin in serum may bind considerable quantities of steroid hormone, but exerts little inhibitory effects on the transport of steroids into brain, whereas globulin-bound hormone does not appear to be transported into brain to a significant extent. Therefore, the hormone fraction in plasma that is available for transport into brain is not restricted to the free (dialyzable) fraction, but includes the larger albumin-bound moiety.