Many estrogenic chemicals found in the environment (xenoestrogens) show a lower affinity for plasma estrogen binding proteins relative to the natural estrogens such as estradiol. These binding proteins, which include alphafetoprotein in rats and mice, sex hormone binding globulin in humans, and albumin in all species, regulate estrogen uptake into tissues. Therefore, the in vivo estrogenic potency relative to estradiol of xenoestrogens that show lower binding to these serum proteins will thus be underestimated in assays that compare the potency of xenoestrogens to estradiol and do not take serum binding into account. We have examined the effects of the binding components in serum on the uptake of a number of xenoestrogens into intact MCF-7 human breast cancer cells. Since most estrogenic chemicals are not available in radiolabeled form, their uptake is determined by competition with [3H]estradiol for binding to estrogen receptors (ER) in an 18-h assay. Serum modified access (SMA) of cell uptake of xenoestrogens is calculated as the RBA in serum-free-medium divided by the RBA in serum, and the bioactive free fraction of xenoestrogen in serum is then also calculated. We predicted the concentration of two xenoestrogens, bisphenol A and octylphenol, required to alter development of the prostate in male mouse fetuses. Whereas octylphenol was predicted to be a more potent estrogen than bisphenol A when tested in serum-free medium, our assay predicted that bisphenol A would be over 500-times more potent than octylphenol in fetal mice. The finding that administration of bisphenol A at a physiologically relevant dose predicted from our in vitro assay to pregnant mice from gestation day 11 to 17 increased adult prostate weight in male offspring relative to controls (similar to the effect of estradiol), while the same doses of octylphenol did not alter prostate development, provided support for our hypothesis.