We examined results from the first national-scale effort to estimate mourning dove (Zenaida macroura) age ratios and developed a simple, efficient, and generalizable methodology for calibrating estimates. Our method predicted age classes of unknown-age wings based on backward projection of molt distributions from fall harvest collections to preseason banding. We estimated 1) the proportion of late-molt individuals in each age class, and 2) the molt rates of juvenile and adult birds. Monte Carlo simulations demonstrated our estimator was minimally biased. We estimated model parameters using 96,811 wings collected from hunters and 42,189 birds banded during preseason from 68 collection blocks in 22 states during the 2005–2007 hunting seasons. We also used estimates to derive a correction factor, based on latitude and longitude of samples, which can be applied to future surveys. We estimated differential vulnerability of age classes to harvest using data from banded birds and applied that to harvest age ratios to estimate population age ratios. Average, uncorrected age ratio of known-age wings for states that allow hunting was 2.25 (SD 0.85) juveniles:adult, and average, corrected ratio was 1.91 (SD 0.68), as determined from harvest age ratios from an independent sample of 41,084 wings collected from random hunters in 2007 and 2008. We used an independent estimate of differential vulnerability to adjust corrected harvest age ratios and estimated the average population age ratio as 1.45 (SD 0.52), a direct measure of recruitment rates. Average annual recruitment rates were highest east of the Mississippi River and in the northwestern United States, with lower rates between. Our results demonstrate a robust methodology for calibrating recruitment estimates for mourning doves and represent the first large-scale estimates of recruitment for the species. Our methods can be used by managers to correct future harvest survey data to generate recruitment estimates for use in formulating harvest management strategies.