We quantified the aggression and feeding of naturally reared steelhead (Oncorhynchus mykiss) fry stocked into a laboratory flume with naturally reared fry or hatchery-reared fry from conventional and enriched rearing environments at three densities in the presence and absence of predators, and compared the aggression and feeding observed in the flume to that observed in two streams. Steelhead fry attack rate increased with density and was reduced in the presence of predators, but was not affected by rearing treatment. Threat rate appeared to increase with density and was significantly affected by rearing treatment combination, but was not significantly affected by predator presence. Feeding rate was not affected by density or rearing treatment, but was reduced in the presence of predators. The rate of aggression by steelhead fry in two streams was lower than that observed in the laboratory and did not increase with density. Rates of aggression and feeding of hatchery-reared and wild steelhead fry were not significantly different in the streams. Overall, we found no evidence that hatchery rearing environments caused higher aggression in steelhead fry. Laboratory observations of salmonid aggression, particularly at high density, may not reflect aggression levels in the wild. ?? 2005 NRC.