The Nebraska Sandhills are an important area for breeding ducks in the Great Plains, but reliable estimates of breeding populations are unavailable. Double-observer methodology was used to estimate abundance of breeding duck populations in the Nebraska Sandhills. Aerial transect surveys were conducted using methodology similar to the cooperative Waterfowl Breeding Population and Habitat Survey conducted by the U.S. Fish and Wildlife Service and Canadian Wildlife Service. Observations by two front-seat observers and one rear-seat observer were used to account for incomplete detectability. Transect-specific population size and detection probabilities were estimated using program SURVIV; estimates were species-specific by type of social grouping. Regional population sizes were obtained by extrapolating transects' estimates to the Sandhills. Detection probabilities were high (>0.75) for all species, but highest for Northern Shoveler (Anas clypeata) and Gadwall (A. strepera). Detection probabilities generally followed increases or decreases with duck densities. Uncorrected population estimates, on average, were <7% the population estimates obtained by correcting for detectability. Double-observer methodology should be considered for adjusting duck counts that cannot be corrected using additional aerial or ground surveys, particularly where water and ducks are well dispersed.