Age classification is important in understanding aspects of population productivity and dynamics, and making informed harvest management decisions. Since 1994 the U.S. Fish and Wildlife Service has, via the Parts Collection Survey, annually collected a wing from a sample of hunter-harvested band-tailed pigeons (Patagioenas fasciata) across the species' range in the United States to estimate productivity over time. Identification of age composition of the harvest depends on a key to quickly and accurately assign pigeons to age classes with as little as a wing from each pigeon. Existing keys for pigeon age classification do not incorporate more recent published information on timing and sequence of post-juvenal molt. We developed a key in 1994 to classify age (juv, subad, ad) of band-tailed pigeons from wings in North America based on earlier published keys (reported to be accurate 97.3%, on average) combined with information subsequently published on timing and sequence of post-juvenal molt. Our key was applied and the examination procedure refined during our annual examining of pigeon wings for the Parts Collection Survey, 1994–2012. We examined all pigeon wings (1/bird) submitted to the Parts Collection Survey. The mean number of wings submitted annually was 494 (range = 114–1,090) from the Pacific Coast population and 79 (range = 3–265) from the Interior population. We examined wings during 1 day each year. Each wing was examined by an initial examiner and final examiner (C. E. Braun). We recorded discrepancies in age classification between examiners during 2005–2012. The mean annual number of discrepancies was 3.1 (SE = 0.3, range = 2–4, n = 8) representing a mean of 1.0% (SE = 0.2, range = 0.4–1.6%) of the total annual sample size. The key has been demonstrated to be accurate and transferable among biologists who have basic training in reading a key and examination of wings (primaries, secondaries, and coverts). Published 2013. This article is a U.S. Government work and is in the public domain in the USA.