It was recently demonstrated that most, if not all, effluents of sewage-treatment works (STWs) in the United Kingdom are estrogenic to fish. As many STWs discharge into rivers, it is possible that some stretches of rivers downstream of where the effluent enters might also be estrogenic. To assess this possibility, the induction of vitellogenin synthesis in caged male trout placed at various distances downstream of the effluent entry point was used as a biomarker of estrogen exposure. Individual discharges into five rivers in England were studied. In four cases, fish placed in the neat effluent, or close to where it entered a river, showed very marked and rapid increases in their plasma vitellogenin concentrations, demonstrating that the effluent was estrogenic. In two of these four cases, none of the downstream sites were estrogenic, whereas in one of the four, fish placed at a site 1.5 km downstream did respond by synthesizing appreciable amounts of vitellogenin, although sites further downstream were not estrogenic. The situation in the fourth river was quite different; not only was the effluent extremely estrogenic (a maximum vitellogenin response in the mg/ml range was attained), but so were all the other study sites on the river, the last of which was 5 km downstream of where the effluent entered. This particular river receives trade effluent from wool-scouring mills, which contains much higher concentrations of alkylphenolic chemicals than any of the other discharges studied. It is suggested that these chemicals probably account for the estrogenic activity of this river. The final (fifth) river showed no estrogenic activity, not even in the neat effluent. This discharge comes from a very small STW, which receives no trade waste, and one or both of these factors may account for why the effluent (and hence the river) was not estrogenic.