Direct emissions from agricultural soils represented approximately 35% (3.28 kt or 1,269 CO2 equivalent kt) of Switzerland’s total nitrous oxide (N2O) emissions in 2022, according to the annual Swiss National Greenhous Gas (GHG) Inventory submitted to the United Nations Framework Convention on Climate Change (UNFCCC). While concerning, estimating N2O emissions over large geographical scales is challenging because of high spatial and temporal variability caused by factors such as climate, soil properties, and land management. This study aims to first improve the Swiss National GHG Inventory by estimating N2O emissions from agricultural soils using the DayCent model in lieu of emission factors. Process-based models have the advantage that they consider the effects of both management and environmental factors on N2O emissions, while emission factors only account for management effects. Second, we compare the resulting, national-scale emissions against estimates based on atmospheric inverse modelling of N2O emissions. For the first part of our study we used newly available, high resolution land use maps to simulate N2O in different land use types (croplands, meadows, and pastures). Simulations were performed based on pedo-climatic conditions that were defined by land use type, regional weather conditions, soil texture, and soil depth. For the second part of the study N2O emissions from this bottom-up approach were compared for a three year period to top-down estimates that were derived from atmospheric observations and transport simulations using inversion techniques. For the years 2021-2023 the bottom-up analysis showed high spatial variability in N2O emissions that could be attributed to both differences in management (e.g. crop types or fertilization intensity) and soil properties. The results of the bottom-up and the top-down approaches were comparable in terms of seasonality. As expected monthly emissions from the top-down approach were generally greater than those of the bottom-up approach as they include other sources of N2O that are not accounted for in the DayCent model (e.g. emissions from manure management and indirect N2O emissions). Similarly, the average percent change in annual N2O emissions from the bottom-up approach was 29% lower than the Swiss National GHG Inventory. Our preliminary results show that atmospheric inversions are very useful to evaluate N2O emission variability at the national scale. The model-based bottom-up approach we set up will allow others to evaluate the effect of different cropping systems under different environmental conditions in future studies and can help stakeholders implement targeted regional measures to reduce emissions.