Nest predation can greatly affect reproductive success of waterfowl. We studied factors influencing duck-nest predation near Humboldt Bay in northern California by testing the prediction that measurements of nest concealment and nest location should differ between successful and depredated artificial nests. We ran a priori hypothesized logistic regression models to evaluate the effects of nest concealment and nest location variables on the fate of all nests, and, for depredated nests, on the type of egg loss. The best-fit model for nest fate included lateral cover and study site variables. Among depredated nests, the best-fit model for the type of egg loss (broken or removed) included vertical cover, study site, and broad habitat. Best-fit models for nest fate and type of egg loss did not include local vegetation, distance to trail, or distance to water. Our results suggest that nest concealment influenced both the likelihood of nest predation and the type of predator to which a nest is vulnerable. We recommend that managers work to maximize opportunities for nest concealment and researchers determine locally important predators in future study designs. TRANSACTIONS OF THE WESTERN SECTION OF THE WILDLIFE SOCIETY 41:11–20; 2005