Abstract Purpose The reduction of the greenhouse gas nitrous oxide (N₂O) to dinitrogen (N₂) via denitrification and N₂O source partitioning between nitrification and denitrification remain major uncertainties in sugarcane systems. We therefore investigated magnitude and product stoichiometry of denitrification and production pathways of N₂O from a tropical sugarcane soil in response to increasing soil nitrate (NO₃⁻) availability. Methods Microcosms were established using a tropical sugarcane soil (Qld, Australia) and emissions of N₂O and N₂ were measured following fertilisation with ¹⁵NO₃⁻–N equivalent to 25, 50 and 100 μg N g⁻¹ soil, simulating soil NO₃⁻ contents previously observed in situ, and mimicking flood irrigation by wetting the soil close to saturation. Results Cumulative N₂O emissions increased exponentially with NO₃⁻ availability, while cumulative N₂ emissions followed an exponential increase to maximum. Average daily N₂ emissions exceeded 5 µg N₂–N g soil⁻¹ and accounted for > 99% of denitrification. The response of N₂O suggests preferential NO₃⁻ reduction with increasing NO₃⁻ availability, increasing N₂O even when NO₃⁻ levels had only a diminishing effect on the overall denitrification rate. The fraction of N₂O emitted from denitrification increased with NO₃⁻ availability, and was a function of soil water, NO₃⁻ and heterotrophic soil respiration. Conclusions Our findings show the exponential increase of N₂O driven by excess NO₃⁻, even though the complete reduction to N₂ dominated denitrification. The low N₂O/(N₂O + N₂) product ratio questions the use of N₂O as proxy for overall denitrification rates, highlighting the need for in-situ N₂ measurements to account for denitrification losses from sugarcane systems.