Global agricultural trade, which increased at the end of 2020, has been described as being “resilient” to the impacts of the COVID-19 coronavirus pandemic; however, the size and channels of its quantitative impacts are not clear. Using a reduced-form, gravity-based econometric model for monthly trade, we estimate the effects of COVID-19 incidence rates, policy restrictions imposed by governments to curb the outbreak, and the de facto reduction in human mobility/lockdown effect on global agricultural trade. We find that while agricultural trade remained quite stable through the pandemic, the sector as a whole did not go unscathed. First, we estimate that COVID-19 reduced agricultural trade by the approximate range of 5 to 10 percent at the aggregate sector level; a quantified impact two to three times smaller in magnitude than our estimated impact on trade occurring in the non-agricultural sector. Reductions in human mobility and policy restrictive responses were the most evident drivers of trade losses. Second, we find sharp differences across individual commodities. In particular, we find that non-food items (hides and skins, ethanol, cotton, and other commodities), meat products including seafood, and higher value agri-food products were most severely impacted by the pandemic; however, the COVID-19 trade effect for the majority of food and bulk agricultural commodity sectors were found to be insignificant, or in a few cases, positive. Third, examining the effect across markets, we find mixed evidence that lower-income and least-developed countries’ trade flows were more sensitive to the pandemic. Fourth, we find evidence that trade flows adjusted to these disruptions over time. Finally, the pandemic also impacted the extensive margins of trade with more severe disruptions detected in air shipments. Findings from this study provide intriguing insights into the dimensions of global agricultural supply chains most resilient and most vulnerable to major global market disruptions.