The binding by serum proteins of circulating thyroid and steroid hormones is a phenomenon whose physiological significance is still not understood, and the validity of the free hormone hypothesis remains in doubt. Indeed, even the most basic physicochemical consequences of these proteins' presence within the microcirculation continue to generate controversy, reflecting disagreement of the rate-limiting effects of hormone-protein interactions on hormone efflux from protein-containing fluid compartments. My colleagues and I have claimed, in particular, that the observations on which Pardridge and coworkers' current ideas crucially depend are entirely explicable using a relatively simple mathematical model of hormone efflux from tissue capillaries differing from the even simpler model relied on by these authors only in that it takes basic hormone-binding kinetics into consideration. The necessity to postulate the local hormone release mechanisms that Pardridge et al. propose in order to account for their observations is thus obviated. Though this conclusion continues to be contested by Pardridge et al. (on the grounds that our own model is invalid), the controversy demonstrates the crucial importance in this area of sound mathematical analysis and the great danger of misinterpreting experimental data by reliance on oversimple theoretical concepts. In reality, the effects of intracapillary protein-binding reactions on target-tissue hormone uptake are of considerably greater complexity than are encompassed in the simple model which is sufficient to explain the observation of Pardridge et al. (1). In particular the assumption made by a number of workers that intracapillary hormone dissociation from binding proteins does not limit the rate of tissue uptake if the latter is substantially less than the intracapillary free hormone generation rate is demonstrably invalid, being incorrectly based on the kinetics of homogeneous (liquid-phase) reactions. By making this assumption, false conclusions may be drawn regarding the kinetics of hormone transport to target tissues, and hence of the effects of the changes in binding protein concentrations that occur in a variety of pathophysiological states. Relying on more detailed analysis, my colleagues and I have suggested that it is plausible, purely on physicochemical grounds, that the characteristic changes in binding protein levels seen in pregnancy serve to redistribute hormone throughout the body, specifically (in the case of the thyroid hormones) directing T4 to the feto-placental unit. Though difficult to verify directly (and perhaps invalid), this proposition has refocused attention on the fetal needs for T4 before the development of the fetal thyroid gland, and on the possible effects on neurological development of an inadequate maternal T4 supply.(ABSTRACT TRUNCATED AT 400 WORDS)