CARBON TRADING and the growing interest in biofuel production from sugarcane necessitate the ability to measure gains and losses of soil organic C which may occur as a result. Modelling and soil sampling suggest that changes in soil C are likely to be < 1 t C/ha/y. Published accounts indicate that confirming such small changes by traditional soil sampling is error-prone and requires investigations of > 10 years. The paper explores the possibility of calculating soil gains or losses by subtracting the carbon stored in the crop biomass from the carbon gained by the crop through the uptake of carbon dioxide supplied by the atmosphere and processes in the soil. Although uptake and storage very nearly balanced each other in one–year measurements in each of two sugarcane crops for which carbon turnover differed by a factor of 2, it was concluded that errors and uncertainties in the measurements and calculations were presently too large to detect the small differences in the gain or loss of soil C claimed for different management practices or predicted by modelling, at least in the short term.