Agricultural soils contribute ~52% of global anthropogenic nitrous oxide (N₂O) emissions, predominantly from nitrogen (N) fertilizer use. Global N₂O emission factors (EFs), estimated using IPCC Tier 1 methodologies, largely rely on short-term field measurements that ignore legacy effects of historic N fertilization. Here we show, through data synthesis and experiments, that EFs increase over time. Historic N addition increases soil N availability, lowers soil pH, and stimulates the abundance of N₂O producing microorganisms and N₂O emissions in control plots, causing underestimates of EFs in short-term experiments. Accounting for this legacy effect, we estimate that global EFs and annual fertilizer-induced N₂O emissions of cropland are 1.9% and 2.1 Tg N₂O-N yr⁻¹, respectively, both ~110% higher than IPCC estimates. Our findings highlight the significance of legacy effects on N₂O emissions, emphasize the importance of long-term experiments for accurate N₂O emission estimates, and underscore the need for mitigation practices to reduce N₂O emissions.

Crop types have significant impacts (p < 0.05) on the background nitrous oxide emissions (BNE) from soils. The annual BNE for China when considering the proportion of planting area is 0.69 ± 1.52 kg N–N₂O ha⁻¹.