EMISSIONS OF the greenhouse gases carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4) from sugarcane crops at Murwillumbah (NSW) and Mackay (Qld) have been measured continuously over whole growing seasons (342 days and 292 days respectively) using chambers and micrometeorological techniques. The second-ratoon crop at Murwillumbah was burnt before harvest, and was grown on an acid sulfate soil (ASS). Green-cane harvesting was practised at Mackay where the crop was in its fifth ratoon. The soil was a Pioneer, non-calcic brown soil. Urea fertiliser was applied to both crops at the start of the measurement period at rates of 160 kgN/ha at Murwillumbah and 150kgN/ha at Mackay. The rates of assimilation of CO2 were 80t/ha at Murwillumbah and 66t/ha at Mackay and were comparable with other tropical crops. At both sites, a large proportion of the CO2 assimilated by the crop came from soil respiration, 35% or 28t/ha at Murwillumbah and 16% or 10t/ha at Mackay. Emissions of N2O at Murwillumbah were large and prolonged, totalling 45.9 kgN/ha over the 342 days and persisting at substantial rates for 5 months. At Mackay, N2O emissions were much smaller totalling 4.7 kgN/ha for the season, most of which was emitted in less than 3 months. Emission factors for N2O were 21% at Murwillumbah and 2.8% at Mackay. The Murwillumbah site was a strong source of CH4 with rates of emission of the same order as those from rice or wetlands, but the net emission over the whole season at Mackay was essentially zero. The sequestering of 51.5 t/ha of CO2 by the crop from the atmosphere at Murwillumbah was partially offset by the emission to the atmosphere of 23 t/ha CO2-e through N2O and CH4. At Mackay, the crop sequestered 55.8 t/ha of CO2, but emitted only 2t/ha CO2-e.