US west coast populations of the native Olympia oyster Ostrea lurida declined precipitously in the late nineteenth and early twentieth centuries and were often replaced by the non-native Pacific oyster (Crassostrea gigas) by the aquaculture industry. Recovery of native oyster ecosystem services derived from their suspension feeding activities (termed “filtration services” (FS)) often serves as a powerful incentive for restoration of populations of O. lurida along the US west coast despite uncertainty about the potential effects of their filtration activities on concentrations of suspended particulate matter. Here, we provide an improved FS model for O. lurida and C. gigas in Yaquina Bay, OR, that is based on both in situ feeding behavior and the complex hydrodynamics of the estuary. The total area and the order of locations chosen for oyster restoration in Yaquina Bay were examined to determine how oyster FS could be maximized with limited resources. These modeling efforts quantified estimates showing (1) native oysters, if restored in Yaquina Bay to historic levels, may contribute nearly an order of magnitude greater FS than previously estimated; (2) C. gigas contributes significantly greater FS than O. lurida, especially during the wet season; (3) FS provided by either species is highly dependent upon seasonal river forcing and salinity; (4) spatial variation in FS arises from the hydrodynamics of the system, uneven oysters distributions, and upstream pre-filtering. We found that spatially explicit models demonstrated the benefits of prioritizing restoration to areas with the greatest FS potential, rather than placing oysters randomly within historic habitats. Directing restoration in this manner used between 75% (dry season) and 60% (wet season) less of the restored area needed to achieve comparable FS with randomly placed oysters.