Rainbow trout are known to be more susceptible to aflatoxin B1 (AFB1) hepatocarcinogenesis than coho salmon, or trout pre-fed the carcinogenesis inhibitors beta-naphthoflavone (beta NF), Aroclor 1254 or indole-3-carbinol. The study reported here examined the relationship between AFB1-glutathione (GSH) conjugation and AFB1 carcinogenesis in salmon, trout and trout pre-fed the three inhibitors. The AFB1-glutathione (AFB1-SG) conjugate was not detected in salmon bile and was present in trout bile in amounts representing less than 0.2% of the administered dose 24 hr after injection of [3H]AFB1. The major conjugates were glucuronides of aflatoxicol and aflatoxicol M1. In incubations of isolated liver cell fractions, less than 0.5% of the original AFB1 dose was recovered as AFB1-SG in salmon and trout preparations, compared to 25% in mouse-liver cell preparations. The GSH concentration in livers of the control trout was higher than that for coho salmon but lower than that for trout pre-fed beta NF. Liver GSH-transferase activity in control trout livers was much higher than in the control salmon livers, but was only 62% of that found for trout fed beta NF. There was no apparent relationship among the various groups between liver GSH concentrations, liver GSH-transferase activity, or biliary GSH conjugate, and the degree of carcinogenic response of AFB1. Thus current evidence does not indicate a major role for aflatoxin B1 epoxide-GSH detoxification in coho salmon, or rainbow trout fed any of the three anticarcinogens tested. These results in salmonid fish are contrary to those which suggest AFB1-SG conjugation as a major determinant of AFB1 carcinogenesis and its dietary modulation in rodent models.