This thesis investigates the sources of hepatic mixed function oxygenase (MFO) enzyme induction and disruptions in circulating hormonal levels in fish exposed to lampricide formulations of 3-trifluoromethyl-4-nitrophenol (TFM). Methodology was developed for the isolation and recovery of chemicals associated with biological activity. Formulation impurities causing MFO induction were isolated using solid phase extraction and HPLC into two fractions. The major constituents in both fractions were identified as nitro-, trifluoromethyl- and/or chloro-substituted diphenyl ethers, which did not elevate MFO activity after in vivo exposures. Three chloro-nitro-trifluoromethyl-substituted dioxin isomers containing these substituents in the most potent fraction were confirmed by characterizations of synthetic isomers. The average concentration of these compounds in previous batches was 288±47 μg/L which translates into loadings of 40 g/year to the Great Lakes Basin. Short-term waterborne exposures to a mixture of 2-trifluoromethyl-3-nitro-7-(and 8)-chloro-dibenzo-p-dioxin isomers elevated MFO activity in rainbow trout (Oncorhynchus mykiss), with a threshold between 0.148-0.745 nM (4.1-20.5 ng/L). Using mammalian (H4IIE) and fish (PLHC-1) in vitro assays, a 2,3,7-substituted isomer was 4-5 times more potent than 2,3,7-trichlorodibenzo-p-dioxin. It was further concluded that the formulation isomers are not substituted exclusively in the 2,3,7-positions. Chemical fractionations directed by competitive binding to rainbow trout estrogen receptors were used in a mechanistically linked approach to identify compounds associated with disruptions in endocrine homeostasis. Virtually all binding potential to the receptor was associated with fractions containing TFM and TFM isomers. Relative to estradiol, TFM demonstrated a competitive affinity of 5.03×10\sp−5. Vitellogenin induction in primary cultures of rainbow trout hepatocytes indicated TFM acts as an estradiol agonist in vitro but protein production was not detected in plasma of caged trout after a lampricide treatment. Following treatment, MFO induction in fish was rapid, peaking within 2-3 d, and was persistent after 18 d in caged rainbow trout, wild white sucker (Catostomus commersoni) and longnose dace (Rhinichthys cataractae) sampled downstream. Induction was highest in fish exposed closest to the application points. Chloro-nitro-trifluoromethyl dioxins were not detected in sediments after treatment and diphenyl ether impurities were <100 ng/g after 18 d.