Managed grazing is the most extensive land use on Earth. The prevailing narrative is that global rangelands, from grasslands to deserts, are being degraded by overgrazing due to overstocking. This perception arises from scientific literature, which contains an order of magnitude more studies on overstocking than on reductions in stocking rates. In contrast, over the past 25 y, regions representing almost half (42%) of global livestock have experienced reductions in stocking rates rather than the expected increases. We evaluated socioeconomic, technological, and climatic direct drivers, as well as indirect drivers, of global stocking patterns. Trade and climate had no detectable effects, whereas technological shifts and meat consumption had an impact on stocking rates. Direct drivers were largely controlled by human population and gross domestic product. Wealthier regions, with slower population growth, greater feed supplementation, and reliance on nongrazing livestock, reduced stocking rates. Less affluent regions, facing rapid population growth and rising meat demand but limited technological and feed resources, increased stocking rates. The overlooked reductions in stocking rates may have major ecological consequences at regional and global scales. Destocking can impact biodiversity, fire regimes, potentially increase carbon sequestration, and modify land-atmosphere fluxes. These effects are not simply the reverse of overstocking, as some consequences are irreversible, including state transitions and local extinctions. The skewed pattern of publications toward overstocking and overgrazing led to a perception of widespread degradation and had consequences on research direction and policy.