The goal of this research was to determine the occurrence of estrogenic activity in source waters, finished drinking waters, and industrial and municipal wastewater effluents. The research team planned to accomplish this goal by validating and optimizing the E-Screen assay, which would then be used to document the estrogenic activity in water samples. Additionally, caged fish studies and in lab exposure studies would be conducted on some of the same samples to assess estrogenic activity with an in vivo system. Using the E-Screen, the research team tested samples from over 70 drinking water facilities. A majority of the source waters (61%) had estrogen activities above the level of detection (0.029 ng/L). Drinking water treatment processes do remove activity; therefore, only 16% of finished waters had activities above the detection limit. Of the 27 wastewater treatment plant effluents tested, activity ranged from no activity to a sample with greater than 1,700 ng/L of activity. Vitellogenin levels were significantly increased in male fathead minnows in only two of the waters tested, both wastewater treatment plant effluents. Using the E-Screen, 90 finished water samples from 72 facilities were tested. Of the finished waters, 84% did not contain estrogenic activity above the level of detection (0.029 ng/L estradiol equivalents). Of the remaining 14 samples, 13 were below 0.27 ng/L. In contrast to the finished waters, the majority (61%) of source waters did have measurable estrogenic activity. Of the 105 surface or groundwater samples tested, 64 were above the detection limit of the assay. Again, the activities were fairly low, with only 10 waters having activities above 0.27 ng/L. Surface waters exhibited higher levels of estrogenic activity than did ground waters; however, 42% of the ground waters tested did have estrogenic activity. Twenty-seven wastewater treatment plant effluents were tested from 17 facilities. Estrogenic activity ranged from no activity (in seven of the effluents) to a sample with greater than 1,700 ng/L of activity. Fish exposure studies were completed at six facilities. No significant responses were found in fish exposed to the source or drinking water samples, but two WWTP effluents caused a significant vitellogenin induction. The low number of positive vitellogenin responses limited their comparison with the E-Screen. Originally published by AwwaRF for its subscribers in 2003.