In the Netherlands, the Dutch Retail Broiler (DRB) and Better Life one Star (BLS) production systems have been introduced with the aim to improve broiler welfare. Simultaneously, retailers set targets for reduction of greenhouse gas (GHG) emissions in the whole broiler production chain. The GHG emissions of DRB and BLS may differ from conventional systems because of differences in slaughter age, feed intake, and diet composition. The aim of this study was to estimate GHG emissions of the conventional, DRB, and BLS production systems. A deterministic, spreadsheet based model was developed that included the breeder, hatchery, and broiler farm stages. First, the model calculates feed intake of different diets and energy use, based on performance objectives and literature. Selection of feed ingredients for the different types of diets was based on least cost formulation with nutritional constraints for each diet. Second, GHG emissions were estimated from cradle to broiler farm gate for processes along the broiler production chain by using life cycle assessment, and expressed as kg CO2-equivalents per kg live weight (kg CO2-eq/kg LW). Results showed that BLS (3.55 kg CO2-eq/kg LW) had lower GHG emissions compared to conventional (3.65 kg CO2-eq/kg LW) and DRB (3.98 kg CO2-eq/kg LW) at the broiler farm gate. Emissions from land use change (LUC) from feed production, mainly from soybean products, had highest impact on total GHG emissions (>50%) for the systems and these soybean products had the lowest inclusion in the diets of the BLS production system. Sensitivity analyses showed that variation in slaughter weight and feed intake could result in overlap of GHG emissions between systems. When soybean products were sourced from a country with low LUC emissions, conventional (1.37 kg CO2-eq/ kg LW) had the lowest GHG emissions and BLS (1.79 kg CO2-eq/kg LW) the highest. This study showed that origin of and including or excluding LUC emissions from soybean production results in different conclusions for achieving the ...