AbstractCrop rotations under no-till (NT) have been a strategy to increase soil organic carbon (SOC) and mitigate greenhouse gas (GHG) emissions, enhancing the cropping system efficiency. However, there is still controversy on the role of grasses and legumes, and species diversity and their impacts. This study aimed to assess the GHG emissions, SOC, and Nitrogen (TN) in a soybean production system managed under NT in rotation with different species in the fall–winter and the spring seasons. Main plots during the fall–winter were (1) Triticale (x Triticosecale) and (2) Sunflower (Helianthus annuus). Subplots established in the spring were (a) Sunn hemp (Crotalaria juncea), (b) Sorghum (Sorghum bicolor), (c) Pearl millet (Pennisetum glaucum), plus a (d) Fallow treatment. Soybean was grown every year in the summer, in sub-subplots. The GHG emission was affected according to crop species. In the spring, Sunn hemp emitted more nitrous oxide (N2O) (0.82 kg ha−1) than fallow (0.58 kg ha−1); however, the high C and N inputs by the legume and also other cover crop residues reduced the relative emissions compared with fallow. Growing pearl millet or Sunn hemp as a spring cover crop increases SOC by 7% on average compared with fallow. The N2O emission of Sunn hemp accounted for only 0.28% of the total N accumulated in the legume residues, notably lower than IPCC estimates. In the fall–winter, Triticale increased SOC by 7%, decreased CO2 emission by 18%, and emitted 20% lower GHG to produce the same soybean yield compared with sunflower. Soybean rotation with triticale in fall–winter and Sunn hemp or pearl millet in spring decreases GHG emissions. Our results indicate that the right choice of species in rotation with soybean under NT increases SOC and may offset GHG emissions from tropical soils. It may be an important tool in mitigating potential global warming.