Abstract: Climate change threatens global food systems, but the extent to which adaptation will reduce losses remains unknown and controversial. Even within the well-studied context of US agriculture, some analyses argue that adaptation will be widespread and damages from climate change will be small, while others conclude that the scope for adaptation is limited and losses will be severe. Globally, scenario-based analyses indicate that, in principle, the extent of adaptation should have major consequences on global agricultural productivity, but there has been no systematic study of how extensively producers actually adapt in the real world at global scale. Here, we empirically estimate the net impact of producer adaptations around the world using longitudinal data on six staple crops spanning 12,658 sub-national units, capturing two-thirds of global crop calories. We project that adaptation and income growth nearly halve global losses at end-of-century, but substantial residual losses remain for all staples except rice. In contrast to analyses of other impact categories that project the greatest damages in poor regions of the world, we find that global damages are dominated by losses to modern-day breadbaskets that currently exhibit limited adaptation due to favorable climates, though losses to many poor regions are also severe. We estimate global production declines 5.8x10^14 kCal annually per 1C rise in global mean surface temperature (4.6% of current production or 130 kCal/person/day, per 1C; p