ABSTRACT-We est~mated the mortality of hatchery-raised pink Oncorhynchus gorbuscha and chum 0 keta salmon fry from seabird predation near a salmon hatchery In Lake Bay, Prince William Sound, Alaska, USA. Fleld counts of seabirds and observations of feeding rates for plunge-diving seabirds were obtained durlng salmon fry releases between Apnl and June 1995. Several hundred birds of 7 piscivorous species aggregated in front of the hatchery. Consumption rates were determined from focalanimal sampling and energetic models. For most species, per capita consumption rates based on behavioral data were lower than those calculated from energetic considerations. From energetic models and fry movement rates, w e estimated that 2.7 to 5.9 million juvenile salmon (1.1 to 2.4 "" of released fry) were consumed during the study period. Differences between the energetics models accounted for 64 % of this range; assumptions about fry movement rates accounted for the remainder. Gulls recorded during aerial surveys throughout Prince William Sound were significantly associated with s p a w n~n g herring and hatchery sites, but not with miles of coast with herring spawn present. The correlation of bird aggregations with the presence of these small fish indicates that concentrations of prey fish have an important influence on the distribution of birds at sea during this time of the year. Bird numbers in both boat counts and aerial surveys declined from early May to early June. Taken together, these results lndlcate that salmon fry just entering the marine environment were not especially susceptible to avlan predation Thelr vulnerability to predation was apparently buffered by the presence of other attractlve food patches within seabird foraging range early in our study period, and by declines in the numbers of seabirds foraging along the shoreline later In the study.