In Australia, pastures form the basis of the extensive livestock industries and are important components of crop rotation systems. Despite recent interest in the soil carbon sequestration value of pastures in the mitigation of climate change, little information is available on the soil carbon sequestration potential of pastures in New South Wales farming systems. To quantify the soil carbon stocks under different pastures and a range of pasture management practices, a field survey of soil carbon stocks was undertaken in 2007 in central and southern NSW as well as north-eastern Victoria, using a paired-site approach. Five comparisons were included: native v. introduced perennial, perennial v. annual, continuous v. rotational grazing, pasture cropping v. control, and improved v. unimproved pastures. Results indicated a wide range of soil organic carbon (SOC) stocks over 0'0.30 m (22.4'66.3 t C/ha), with little difference when calculated based on either constant soil depth or constant soil mass. Significantly higher SOC stocks were found only as a result of pasture improvement using P application compared with unimproved pastures. In this case, rates of sequestration were estimated to range between 0.26 and 0.72 t C/ha.year, with a mean rate of 0.41 t C/ha.year. Lack of significant differences in SOC stocks for the other pastures and pasture management practice comparisons could be due to inherent problems associated with the paired-site survey approach, i.e. large variability, difficulties in obtaining accurate site history, and the occasional absence of a valid control as well as the likely lower rates of SOC sequestration for these other comparisons. There is a need for scientific long-term trials to quantify the SOC sequestration potential of these other pastures and pasture management practices.