We used an experimental aerial transect survey with stratified random sampling, to estimate the size of the population of wintering black ducks (Anas rubripes) in coastal New Jersey during 2 winters, and the coastal Atlantic Flyway (Me. to S.C.) during 4 years. Population estimates were precise (CV < 0.20) on a flyway basis, whereas individual strata (states) had coefficients of variation of 0.16-0.71. Population estimates agreed with the conventional mid-winter waterfowl surveys (MWS) for all 4 years (MWS within 95% CI of N). We recommend continuing these surveys to provide precise and statistically defensible population estimates for black ducks. Additional improvements in precision may be achieved using recent developments in estimation such as Bayesian techniques. Techniques to decrease bias through air-ground comparisons are likely to be expensive and will require more development. Air-ground comparisons can probably be justified if there is a demonstrable need for an estimate of the absolute size of the black duck population versus an index.