To more effectively use ospreys as a biomonitoring tool and to better assess contaminant pathways, the diet of nesting ospreys (Pandion haliaetus) was studied along the lower Columbia and upper mainstem Willamette rivers by evaluating prey remains collected from wire baskets constructed under artificial feeding perches installed near nest sites and from the ground beneath natural feeding perches and nests. Prey remains from 1997-2004 on the Columbia River and 1993 (previously published) and 2001 on the Willamette River were evaluated and compared. Largescale suckers (Catostomus macrocheilus) were the predominate fish species identified in collections from the Columbia River (61.5% [84.3% biomass]) and Willamette River (76.0% [92.7% biomass]). Prey fish diversity, when based only on ground collections, was higher in the Columbia (2.45) than the Willamette river (1.92) (P = 0.038). Prey fish diversity in collections from the Willamette River did not differ between this study (2001) and previous study (1993) (P = 0.62). Fishbones recovered in wire baskets are likely more representative of osprey diet compared to bones recovered from the ground, because prey diversity was higher among basket samples compared to ground collections (wire basket diversity = 5.25 vs. ground collection diversity = 2.45, P = 0.011). Soft-boned salmonids (Oncorhynchus spp.), American shad (Alosa sapidissima), and mountain whitefish (Prosopium williamsoni) were probably underrepresented in collections obtained from the ground. Study results suggest that baskets provide a better method for assessing osprey diet than other indirect methods. These findings augment available osprey food-habits information and provide additional biological and ecological information to better assess potential impacts of various environmental contaminants on nesting ospreys.