Two kinds of estradiol binding sites are present in purified nuclei from the rat uterus following estradiol injection. One of these sites (type I) corresponds to the well-known estrogen receptor which undergoes translocation from the cytoplasm to the nucleus. The second site (type II) is not translocated from the cytoplasm to the nucleus, however, estradiol treatment does stimulate an increased number of these sites. Type II sites are observed in purified nuclei and chromatin isolated from the uterus but not from non-target tissues such as the spleen and diaphragm. Thus an elevation in the levels of type II sites appear to be a specific nuclear response of the rat uterus to estradiol. Saturation analysis over a wide range of [3H]-estradiol concentrations produces a binding curve for type II sites which is sigmoidal and hence no accurate estimation of the dissociation constant is possible. The binding of [3H]-estradiol to nuclear type II sites is inhibited by estradiol and diethylstilbestrol but not by progesterone, testosterone, or corticosterone. Extraction of nuclei isolated from estrogen treated rat uteri with KCl provides a complex picture. Direct labeling of nuclear estrogen receptors either by in vivo injection or in vitro incubation of intact uteri with [3H]-estradiol measures only a fraction of the specific estrogen binding sites associated with the nuclear pellet following 0.4 M KCl extraction. These sites are more accurately determined by performing saturation analysis over a wide range of [3H]-estradiol concentrations by exchange which measures specific estrogen binding sites, not [3H]-steroid. Saturation analysis of estradiol binding to KCl extracted nuclei when performed by exchange, with appropriate corrections for type II binding, reveals that approximately 1000--2000 receptors per nucleus are resistant to KCl extraction 1 hr after administration. The same numbers of type I sites display long-term nuclear retention. A single injection of estradiol results in long term (greater than 6 h) retention of type I sites, rapid and sustained elevations (1--72h) in type II sites and true uterine growth (uterine wet weight at 24--43 h). Estriol injections caused a rapid increase in nuclear type I sites which was not accompanied by an increase in type II sites and no true uterine growth occurred. Administration of estriol or estradiol as a pellet implant, which causes continuous occupancy of type I sites, increases the quantity of nuclear type II sites and stimulates true uterine growth. Therefore, we conclude that elevated levels of nuclear type II sites correlate with the long term uterotropic response to estrogenic hormones. Although we do not understand the function of this second class of binding sites it is possible that the type II sites represent a major component in the mechanism by which estrogens stimulate growth of the uterus.