Prescribed wildland and agricultural fires are common in the eastern United States (U.S.), but their small size and short durations have led polar-orbiting satellite datasets to underestimate their emissions due to omission. The higher temporal resolution of the GOES-16 Advanced Baseline Imager (ABI) improves fire detection, enabling development of a new biomass burning emissions inventory: GOES Eastern U.S. Fire Emissions (GEUFE, pronounced “goofy”). GEUFE is based on geostationary fire radiative power (FRP) observations, incorporates improved crop type classification for agricultural fires, and includes a new emission factor (EF) compilation for 16 land cover types in the eastern U.S., including forest, crop, and grassland types that differ markedly from globally reported EF values. GEUFE provides daily estimates of dry matter (DM) consumption and emissions of CO2, CO, NOx, NH3, SO2, organic carbon (OC), black carbon (BC), and fine particulate matter (PM2.5) at 2 km resolution. An intercomparison of nine biomass burning emissions inventories (GEUFE, GBBEPx, QFED, NEI-2017, FEER, VFEI, GFAS, FINN, GFED) showed that OC estimates in the eastern U.S. during August 2019–July 2020 ranged from 0.05 to 1.32 Tg yr-1, with both EFs and DM consumption driving the large spread in aerosol estimates. GEUFE, at 0.64 Tg yr-1, is at the high end of this range due both to region-specific EFs and continuous GOES-16 monitoring, reducing missed short-lived fires. GEUFE is a process-based inventory that eliminates reliance on empirical scaling factors, such as QFED’s aerosol optical depth-based adjustments, while explaining previously inferred “missing emissions” from small fires.