Cougar (Puma concolor) management has been hindered by inability to identify population trends. We documented changes in sex and age of harvested cougars during an experimentally induced reduction in population size and subsequent recovery to better understand the relationship between sex-age composition and population trend in exploited populations. The cougar population in the Snowy Range, southeast Wyoming, was reduced by increased harvest (treatment phase) from 58 independent cougars (>1 year old) (90% CI = 36-81) in the autumn of 1998 to 20 by the spring of 2000 (mean exploitation rate = 43%) and then increased to 46 by spring 2003 following 3 years of reduced harvests (mean exploitation rate = 18%). Pretreatment harvest composition was 63% subadults (1.0-2.5 years old), 23% adult males, and 14% adult females (2 seasons; n = 22). A reduction in subadult harvest, an initial increase followed by a reduction in adult male harvest, and a steady increase in adult female harvest characterized harvest composition trends during the treatment phase. Harvest composition was similar at high and low densities when harvest was light, but proportion of harvested subadult males increased at low density as they replaced adult males removed during the treatment period (high harvest). While sex ratio of harvested cougars alone appears of limited value in identifying population change, when combined with age class the 2 appear to provide an index to population change. Composition of the harvest can be applied to adaptively manage cougar populations where adequate sex and age data are collected from harvested animals.