In vitro estrogenicity screens are believed to provide a first prioritization step in hazard characterization of endocrine disrupting chemicals. When applied to complex environmental matrices or mixture samples, they have been indicated valuable in estimating the overall estrogen-mimicking load. In this study, the performance of an adapted format of the classical E-screen or MCF-7 cell proliferation assay was profoundly evaluated to rank pure compounds as well as influents and effluents of sewage treatment plants (STPs) according to estrogenic activity. In this adapted format, flow cytometric cell cycle analysis was used to allow evaluation of the MCF-7 cell proliferative effects after only 24h of exposure. With an average EC value of 2pM and CV of 22%, this assay appears as a sensitive and reproducible system for evaluation of estrogenic activity. Moreover, estrogenic responses of 17 pure compounds corresponded well, qualitatively and quantitatively, with other in vitro and in vivo estrogenicity screens, such as the classical E-screen (R=0.98), the estrogen receptor (ER) binding (R=0.84) and the ER transcription activation assay (R=0.87). To evaluate the applicability of this assay for complex samples, influents and effluents of 10 STPs covering different treatment processes, were compared and ranked according to estrogenic removal efficiencies. Activated sludge treatment with phosphorus and nitrogen removal appeared most effective in eliminating estrogenic activity, followed by activated sludge, lagoon and filter bed. This is well in agreement with previous findings based on chemical analysis or biological activity screens. Moreover, ER blocking experiments indicated that cell proliferative responses were mainly ER mediated, illustrating that the complexity of the end point, cell proliferation, compared to other ER screens, does not hamper the interpretation of the results. Therefore, this study, among other E-screen studies, supports the use of MCF-7 cell proliferation as estrogenicity screen for pure compounds and complex samples.