Problems contributing to the low fish guidance at Rocky Reach Dam were investigatedthrough laboratory experiments and a computer model. Particular attention was given todetermining how strobe lights could be used to improve fish guidance efficiency.In general, juvenile Chinook and coho salmon and rainbow trout tended to avoid or escape strobelight at intensities between 0.1 and 5 ìE/m2/s. The EG&G strobe lights used at Rocky ReachDam produced an intensity of about 1 ìE/m2/s at a distance of 10 ft. Escape speed averagedabout 7 bI/s and was dependent on strobe light intensity and the level of light that fish wereadapted to. Escape speed increased with strobe flash rate, with a flash rate of 200/mm or abovebeing sufficient to induce speed near the maximum observed. Strobe intensity above 5 ìE/m2/sand flash rates above about 500/mm tended to stun fish. Juvenile steelhead were attracted to thelow light intensities of the strobe's penumbra.A computer model was developed to describe the trajectories of fish as affected by water currentsand swimming to avoid strobe lights and the trashrack. A model suggested that trashrackpassage was a critical factor in fish guidance efficiency. In laboratory experiments, strobe lightcould be used to force rainbow trout through a trashrack-like barrier. However, at Rocky ReachDam, strobe lights were ineffective at forcing subyearling Chinook through the trashrack. Thus, ahypothesis was formulated that the submersible traveling screen behind the trashrack may havegenerated sound stimuli that inhibited trashrack passage; moreover, it appeared that strobe lightswere ineffective in counteracting this aversive stimuli.