1) In our previous paper, it was reported that in human endometrium the step of formation of nuclear receptor -estradiol complex was not temperature dependent but in rat uteri temperature dependent. In order to clarify whether this temperature dependency in rat uteri means an energy requirement for this step or not, experiments with 2,4 dinitrophenol as an energy uncoupler were done. Even when rat uteri or endometrium of women were incubated with 10 muc of 3H-estradiol and 10(-4)M of 2,4-dinitrophenol, 8S estrogen receptor-E2 complex in cytosol and 5S estrogen receptor-E2 complex in nuclear extract were recognized in both rat and woman. Namely, no evidence was recognized for this step to require activation energy in rat uterus as well as in human endometrium. 2) The cytosol fraction was prepared after human endometrium had been incubated with 3HE2 at 2 degrees C. Cytosol was analysed on sucross density gradient after heat treatment (at 10 degrees C, 31 degrees C or 37 degrees C). At 31 degrees C and 37 degrees C treatment, the 8S receptor peak was difficultly recognized. Namely, the 8S receptor-E2 complex was heat labile in cell free condition. This appears to be one of the reasons why 8S cytosol receptor was not visible in human endometrium incubated with 3HE2 at 37 degrees C. 3) When the cytosol fraction from the human endometrium was analysed on 3--20% linear sucrose gradient containing 0.4M KCl, the 8S peak diminished and about 4S peak appeared. This was considered to mean that 8S receptor in human endometrium was also split to a 4S binding unit at high salt condition and the 8S receptor in humans had also a subunit structure. 4) 8S receptor in cytosol and 5S receptor in nuclear fraction were recognized in one case of human endometrial carcinoma, but in another case both receptors were not recognized. 5) Estrogen binding protein in human serum was sedimented at 6S fraction by sucrose gradient.