We used life cycle assessment methodology to determine the cradle-to-farmgate GHG emissions for rainfed wheat grown in monoculture or in sequence with the break crops canola (Brassica napus) and field peas (Pisum sativum), and for the break crops, in the south-eastern grains region of Australia. Total GHG emissions were 225 kg carbon dioxide equivalents (CO₂-e)/t grain for a 3 t/ha wheat crop following wheat, compared with 199 and 172 kg CO₂-e/t for wheat following canola and field peas, respectively. On an area basis, calculated emissions were 676, 677 and 586 kg CO₂-e/ha for wheat following wheat, canola and field peas, respectively. Highest emissions were associated with the production and transport of fertilisers (23–28% of total GHG emissions) and their use in the field (16–23% of total GHG emissions). Production, transport and use of lime accounted for an additional 19–21% of total GHG emissions. The lower emissions for wheat after break crops were associated with higher yields, improved use of fertiliser nitrogen (N) and reduced fertiliser N inputs in the case of wheat after field peas. Emissions of GHG for the production and harvesting of canola were calculated at 841 kg CO₂-e/ha, equivalent to 420 kg CO₂-e/t grain. Those of field peas were 530 kg CO₂-e/ha, equivalent to 294 kg CO₂-e/t grain. When the gross margin returns for the crops were considered together with their GHG emissions, the field pea–wheat sequence had the highest value per unit emissions, at AU$787/t CO₂-e, followed by wheat–wheat ($703/t CO₂-e) and canola–wheat ($696/t CO₂-e). Uncertainties associated with emissions factor values for fertiliser N, legume-fixed N and mineralised soil organic matter N are discussed, together with the potentially high C cost of legume N₂ fixation and the impact of relatively small changes in soil C during grain cropping either to offset all or most pre- and on-farm GHG emissions or to add to them.