The greenhouse gas fluxes and effective mitigation strategies in subtropical vegetable cropping systems remain unclear. In this field experiment, nitrous oxide (N2O) and methane (CH4) fluxes from an irrigated lettuce cropping system in subtropical Queensland, Australia, were measured using manual sampling chambers. Four treatments were included: Control (no fertiliser), U100 (100 kg N ha–1 as urea), U200 (200 kg N ha–1 as urea) and N100 (100 kg N ha–1 as nitrate-based fertilisers). The N fertilisers were applied in three splits and irrigation was delivered sparingly and frequently to keep soil moisture around the field capacity. The cumulative N2O emissions from the control, U100, U200 and N100 treatments over the 68-day cropping season were 30, 151, 206 and 68 g N2O-N ha–1, respectively. Methane emission and uptake were negligible. Using N2O emission from the Control treatment as the background emission, direct emission factors for U100, U200 and N100 treatments were 0.12%, 0.09% and 0.04% of applied fertiliser N, respectively. Soil ammonium (NH4+) concentration, instead of nitrate (NO3–) concentration, exhibited a significant correlation with N2O emissions at the site where the soil moisture was controlled within 50%–64% water-filled pore space. Furthermore, soil temperature rather than water content was the main regulating factor of N2O fluxes in the fertilised treatments. Fertiliser type and application rates had no significant effects on yield parameters. Partial N balance analysis indicated that approximately 80% and 52% of fertiliser N was recovered in plants and soil in the treatments receiving 100 kg N ha–1 and 200 kg N ha–1, respectively. Therefore, in combination with frequent and low-intensity irrigation and split application of fertiliser N, substitution of NO3–-based fertilisers for urea and reduction in fertiliser N application rates were considered promising mitigation strategies to maintain yield and minimise N2O emissions during the low rainfall season.