Previous efforts to relate winter-ground precipitation to subsequent reproductive success as measured by the ratio of juveniles to adults in the autumn failed to account for increased vulnerability of juvenile ducks to hunting and uncertainty in the estimated age ratio. Neglecting increased juvenile vulnerability will positively bias the mean productivity estimate, and neglecting increased vulnerability and estimation uncertainty will positively bias the year-to-year variance in productivity because raw age ratios are the product of sampling variation, the year-specific vulnerability, and year-specific reproductive success. Therefore, we estimated the effects of cumulative winter precipitation in the California Central Valley and the Mississippi Alluvial Valley on pintail (Anas acuta) and mallard (Anas platyrhnchos) reproduction, respectively, using hierarchical Bayesian methods to correct for sampling bias in productivity estimates and observation error in covariates. We applied the model to a hunter-collected parts survey implemented by the United States Fish and Wildlife Service and band recoveries reported to the United States Geological Survey Bird Banding Laboratory using data from 1961 to 2013. We compared our results to previous estimates that used simple linear regression on uncorrected age ratios from a smaller subset of years in pintail (1961–1985). Like previous analyses, we found large and consistent effects of population size and wetland conditions in prairie Canada on mallard productivity, and large effects of population size and mean latitude of the observed breeding population on pintail productivity. Unlike previous analyses, we report a large amount of uncertainty in the estimated effects of wintering-ground precipitation on pintail and mallard productivity, with considerable uncertainty in the sign of the estimated main effect, although the posterior medians of precipitation effects were consistent with past studies. We found more consistent estimates in the sign of an interaction effect between population size and precipitation, suggesting that wintering-ground precipitation has a larger effect in years of high population size, especially for pintail. When we used the estimated effects in a population model to derive a sustainable harvest and population size projection (i.e., a yield curve), there was considerable uncertainty in the effect of increased or decreased wintering-ground precipitation on sustainable harvest potential and population size. These results suggest that the mechanism of cross-seasonal effects between winter habitat and reproduction in ducks occurs through a reduction in the strength of density dependence in years of above-average wintering-ground precipitation. We suggest additional investigation of the underlying mechanisms and that habitat managers and decision-makers consider the level of uncertainty in these estimates when attempting to integrate habitat management and harvest management decisions. Collection of annual data on the status of wintering-ground habitat in a rigorous sampling framework would likely be the most direct way to improve understanding of mechanisms and inform management.