Estrogens stimulate growth of MCF-7 breast cancer cells in monolayer culture. Possible interference of serum factors leading to an estrogen-insensitive cell growth was analyzed in various experiments carried out on serum batches producing no estradiol stimulation. Out of five estrogen conjugates, only 3-glucurono-estradiol partly suppressed the inhibition of hydroxytamoxifen; the conjugate also reduced the estrogen receptor content of the cells, probably by a down regulation process ("processing"). Moreover, prolonged subcultures in dextran-coated charcoal-treated serum attempting to remove possible intracellular estrogens produced no growth stimulation. Interference by hormone carriers of the serum was ruled out by the fact that two strong synthetic estrogens, moxestrol and diethylstilbestrol with weak binding affinity for these carriers, were unstimulatory. Reduction of the carrier concentration also failed to confer any estrogen sensitivity. This lack of effect of most estrogen conjugates and serum carriers seems to contradict the hypothesis of their interference leading to an estrogen-insensitive growth. Presence in the serum of potential inhibitors towards estrogen action was also examined. Dilution of sera inducing an estrogenic stimulated growth failed to show any growth increase, either in the absence or presence of estradiol, thus excluding the possibility of a major influence of an antagonism on growth control. Moreover, clonogenic assays in soft agar eliminated the hypothesis that a difference between "active" (stimulatory with estradiol) and "inactive" serum batches may result from distinct adherence properties rather than from real growth stimulation. All of these data are consistent with the concept that serum factors which are not of estrogenic nature mediate the trophic effect of estradiol; their absence in some serum batches may lead to an estrogen-insensitive cell growth.