The value of riparian zones to nongame birds and islands to nesting waterfowl has been well documented. Therefore, we predicted that nongame birds would have higher densities and nest success on riverine islands because they would be better protected from predators. Furthermore, because larger islands may support resident mammalian predators, we hypothesized that nest success would decrease with increasing island size. In 1998 and 1999, we compared breeding bird density, nest success, habitat, and potential nest predators between islands (n=44) and mainland sites (n=44) on the middle Snake River in Idaho and Oregon, USA. Ground-, shrub-, and tree-nesting guilds had higher densities on islands, while island was a significant predictor in 5 of 10 species-habitat models. Despite this trend, neither daily survival rates of nests by nesting guilds or proportion of riparian cover types were different between islands and mainland sites. We also could detect no relationship between nest-survival rates and island size. Our camera study on artificial nests demonstrated that nongame-bird nests were available to a wider range of predators than those typically identified in waterfowl studies. Nest predators such as deer mice (Permomyscus maniculatus) and squirrels (Sciurus spp.) often were resident on even the smallest islands. Black-billed magpies (Pica hudsonia) likely were a major nest predator along the Snake River, where they were found in extremely high densities. We also observed a trend toward higher brown-headed cowbird (Molothrus ater) densities and parasitism rates on islands. Influence of landscape-level factors, such as river morphology and land use, on the distribution and abundance of nongame birds and nest predators requires additional research. Our research demonstrated that agencies managing islands for nesting waterfowl should not assume that those same islands also provide refuge for nongame birds.