Abstract Studies at the Pacific Northwest Laboratory have evaluated fish behavior and migration in response to gas-supersaturated water, thermal discharge, water-soluble fractions (WSFs) of coal liquids and other environmental stresses. Approaches have included biotelemetry in the field, and avoidance/attraction and predator/prey studies in the laboratory. This article specifically addresses three study examples and integrates the results with those of related studies. Overall, major findings included the following: Thermal discharges (surface water Δts 0->17°C) did not block upstream migration of sonic-tagged adult chinook salmon (Oncorhynchus tschawytscha) and rainbow trout (O. mykiss, formerly Salmo gairdneri) in the Hanford Reach of the Columbia River. Juvenile chinook salmon avoided thermal discharges in the laboratory when Δts exceeded 9 to 11°C above ambient. However, juvenile salmon were more susceptible to predation at 10 to 20% of the thermal dose causing loss of equilibrium. Radio-tagged adult chinook salmon swam deeper in supersaturated water than in normally saturated water in the Snake River and, thereby, avoided the upper, critical zone. Carp (Cyprinus carpio) and black bullhead (Ictalurus melas) did not always avoid lethal gas levels in the laboratory and some fish died in the test apparatus. Fathead minnow (Pimephales promelas) avoided the WSF of a coal liquid at concentrations causing acute effects but not at those causing chronic effects. Rainbow trout did not avoid coal liquid WSFs although they reportedly avoid the major constituent, phenol, tested as a pure compound. Susceptibility to predation of juvenile rainbow trout did not increase until phenol concentrations reached the acute LC50. A conceptual model to link avoidance and toxicological data for environmental assessment is presented.