European Union (EU) Member States have agreed to limit their greenhouse gas (GHG) emissions from sectors not covered by the EU Emissions Trading Scheme (non-ETS). That includes also emissions from agricultural sector. Although the Intergovernmental Panel on Climate Change (IPCC) has established a methodology for assessment of GHG emissions from agriculture, the forecasting options are limited, especially when policies and their interaction with the agricultural system are tested. Therefore, an advanced tool, a system dynamics model, was developed that enables assessment of effects various decisions and measures have on agricultural GHG emissions. The model is based on the IPCC guidelines and includes the main elements of an agricultural system, i.e. land management, livestock farming, soil fertilization and crop production, as well as feedback mechanisms between the elements. The case of Latvia is selected for simulations, as agriculture generates 22% of the total anthropogenic GHG emissions in the country. The results demonstrate that there are very limited options for GHG mitigation in the agricultural sector. Thereby, reaching the non-ETS GHG emission targets will be very challenging for Latvia, as the level of agricultural GHG emissions will be exceeded considerably above the target levels. Thus, other non-ETS sectors will have to reduce their emissions drastically to “neutralize” the agricultural sector’s emissions for reaching the EU’s common ambition to move towards low-carbon economy. The developed model may serve as a decision support tool for impact assessment of various measures and decisions on the agricultural system’s GHG emissions. Although the model is applied to the case of Latvia, the elements and structure of the model developed are similar to agricultural systems in many countries. By changing numeric values of certain parameters, the model can be applied to analyze decisions and measures in other countries.