Many assessments of product carbon footprint (PCF) for agricultural products omit emissions arising from land-use change (LUC). In this study, we developed a framework based on IPCC national greenhouse gas inventory methodologies to assess the impacts of LUC from crop production using oil palm, soybean and oilseed rape as examples. Using ecological zone, climate and soil types from the top 20 producing countries, calculated emissions for transitions from natural vegetation to cropland on mineral soils under typical management ranged from −4.5 to 29.4 t CO2-eq ha−1 yr−1 over 20 years for oil palm and 1.2–47.5 t CO2-eq ha−1 yr−1 over 20 years for soybeans. Oilseed rape showed similar results to soybeans, but with lower maximum values because it is mainly grown in areas with lower C stocks. GHG emissions from other land-use transitions were between 62% and 95% lower than those from natural vegetation for the arable crops, while conversions to oil palm were a sink for C. LUC emissions were considered on a national basis and also expressed per-tonne-of-oil-produced. Weighted global averages indicate that, depending on the land-use transition, oil crop production on newly converted land contributes between −3.1 and 7.0 t CO2-eq t oil production−1 yr−1 for palm oil, 11.9–50.6 t CO2-eq t oil production−1 yr−1 for soybean oil, and 7.7–31.4 t CO2-eq t oil production−1 yr−1 for rapeseed oil. Assumptions made about crop and LUC distribution within countries contributed up to 66% error around the global averages for natural vegetation conversions. Uncertainty around biomass and soil C stocks were also examined. Finer resolution data and information (particularly on land management and yield) could improve reliability of the estimates but the framework can be used in all global regions and represents an important step forward for including LUC emissions in PCFs.