In this study, the Organisation for EconomicCooperationand Development (OECD) Test Guideline 204 was adapted for the development of an in vivo screen for detecting estrogenic activity using juvenile rainbow trout (Oncorhynchus mykiss). The influence of feeding ration and the effect of reference estrogenic chemicals on somatic growth, gonadosomatic index, hepa-tosomatic index, and plasma vitellogenin concentrations were evaluated over 21 d. A feeding ration of 4% body weight per day resulted in temporal increases in body weight, hepatosomatic index, and plasma vitellogenin concentration in both males and females. A feeding ration of 1% did not induce any changes in the measured endpoints and was adopted for the subsequent estrogen exposures. Exposure of juvenile female fish to 17β-estradiol and 4-tert-nonylphenol produced concentration-dependent inductions of plasma vitellogenin that were optimal after 14 d, with lowest observed effect concentrations of 8.9 ng/L and 16 μg/L, respectively. Methoxychlor also induced synthesis of plasma vitellogenin with a 14-d lowest observed effect concentration of 9.8 μg/L. Higher concentrations of methoxychlor were toxic to fish. Increases in the hepatosomatic index occurred on day 14 in fish exposed to 244 ng/L 17β-estradiol and 53μg/L 4-tert-nonylphenol. None of the compounds tested caused detectable changes in the gonadosomatic index. Vitellogenin induction in juvenile trout was sensitive to estrogenic chemicals, occurred rapidly, and the response was highly consistent between experiments (coefficient of variation = 4.5% for the 17β-estradiol positive control). The 14-d juvenile fish assay offers a sensitive and robust in vivo screening method for estrogen-active substances.