Maintenance of soil carbon (C) content is important because a relatively small percentage change in the global soil C store has the potential to cause a large change in atmospheric CO₂ concentration. Losses of soil C can also lead to a decline in soil quality and its capacity to be productive and carry out other services such as the filtering of pollutants. Globally, research on soil C dynamics has largely focused on forests, croplands and natural grasslands, while intensively grazed pasture has received much less attention. In New Zealand, the dynamics of soil C content and C cycling in intensively grazed dairy systems are poorly understood but large losses of soil C (1 t C ha⁻¹ yr⁻¹) have recently been reported for grazed dairy pastures. The objective of this research was to build on current knowledge of the C balance of intensively grazed dairy farm systems. To achieve this objective, net ecosystem CO₂ exchange (NEE) and water use efficiency (WUE) were measured over intensively grazed dairy pasture using eddy covariance from 15 December 2007 to 15 December 2009. Net ecosystem carbon balances (NECB) were then calculated for 2008 & 2009 from NEE measurements combined with measurements and estimates of C imports (feed) and C exports (milk, silage, methane). A further objective was to determine the impact of periodic cultivation of contrasting soils on the C balance of a dairy farm. To achieve this objective, measurements of soil CO₂ emissions were made using the closed chamber technique following the cultivation of three paddocks of Horotiu soil (Typic Orthic Allophanic) and three paddocks of Te Kowhai soil (Typic Orthic Gley). Annual NEE of the farm was -1,212 ± 500 kg C ha⁻¹ for 2008 and -2,280 ± 500 kg C ha⁻¹ for 2009. Including imports and exports of C to the farm resulted in an annual NECB of -199 ± 500 kg C ha⁻¹ and -1,014 ± 500 kg C ha⁻¹ for 2008 and 2009, respectively. Applied uncertainty is at 90% confidence bound and derived from previous studies reported in the literature. The site was a net sink of C during both 2008 and 2009 in agreement with EC studies performed over grasslands in Europe. The large difference in NEE and NECB between years was due to a drought in 2008, when the site was a C source for the first four months of this year. Average daily water use efficiency (WUE) for 2008 (4 g C kg⁻¹H₂O) and 2009 (4.2 g C kg⁻¹ H₂O) were not substantially different between years and agreed with international field and laboratory studies for pasture. Soil CO₂ loss following cultivation was measured using the closed chamber technique. During the period of cultivation photosynthesis ceased, and potential C input (NEE) to pasture during this time was estimated at -750 kg C ha⁻¹from the adjacent EC study site. To calculate the maximum net soil CO₂₋C loss the potential C input from photosynthesis (NEE) must be added to measured CO₂ emissions. Total soil C loss from the Te Kowhai was between 2,880 kg C ha⁻¹ (CO₂ flux only) and 3,742 kg C ha⁻¹ (CO₂ flux + NEE) while the Horotiu soil lost between 2,082 kg C ha⁻¹(CO₂ flux only) and 2,944 kg C ha⁻¹ (CO₂ flux + NEE). The significant difference in C loss between the two soils was likely a result of their contrasting clay mineralogy and drainage. The Horotiu soil contains allophanic clays with a very high specific surface area, which protects soil C from decomposition. Additionally, poorly drained soils such as the Te Kowhai tend to lose more C following cultivation due to aeration caused by cultivation which increases oxygen penetration into the soil and accelerates decomposition of soil C. Based on these results this grazed pasture was a net sink of C for 2008 and 2009 which is in contrast to the measured decline of 1 t C ha⁻¹ yr⁻¹ from New Zealand’s flat to rolling dairy pastures. Cultivation of dairy pasture soil resulted in net C losses, however, these losses were not large enough to account for the measured decline in soil C from New Zealand’s flat to rolling dairy pastures. Further research is required to investigate long term soil C recovery following initial cultivation of pasture in order to be confident of this conclusion.