Rainbow trout metabolized benzene to phenol and catechol, and excreted them primarily as glucuronide conjugates. Treatment with benzene or toluene resulted in spleenomegaly that may be a more generalized solvent effect on the circulating RBCs. Anemia was observed only in the benzene-treated animals. This may be a result of decreased erythropoiesis in the head kidney or a direct effect of the principal metabolite (phenol) or its reactive intermediate. Both the spleen and liver metabolized benzene in vitro, while no metabolism was observed in the head kidney or trunk kidney. This fact, along with the observed "covalently" bound material in the liver, spleen, and kidney in vivo, can be interpreted as indicating that a reactive metabolite is formed outside the head kidney and is transported to the head kidney where it results in the observed hematopoietic damage. A similar hypothesis has been suggested for the observed protection by partial hepatectomy against bone marrow effects in mammalian species. The effects observed in trout are similar to those reported for higher mammalian species, but the metabolic profile is simpler, and studies in these animals may shed light on the mechanism of benzene toxicity. The techniques developed for monitoring benzene toxicity in trout can also be used for examining other environmental pollutants that may affect the hematopoietic system of fish.