Grazed grasslands occupy 26% of the earth's ice free land surface and are therefore an important component of the global C balance. In New Zealand, pastoral agriculture is the dominant land use and recent research has shown that soils under intensive dairy pastures have lost large amounts of carbon (∼1000 kg C ha⁻¹ y⁻¹) during the past few decades. The objective of this research was to determine the net ecosystem carbon balance (NECB) of an intensively grazed dairy pasture in New Zealand. Net ecosystem CO₂ exchange (NEE) was measured using an eddy covariance (EC) system from 1 January 2008 to 31 December 2009. Other C imports (feed) and exports (milk, methane, leaching, and harvested biomass) were calculated from farm production data and literature values. During 2008 there was a one in 100 year drought during summer/autumn, which was followed by a very wet winter. There were no prolonged periods of above or below average rainfall or soil moisture in 2009, but temperatures were consistently lower than 2008. The severe summer/autumn drought during 2008 caused a loss of CO₂ to the atmosphere, but annual NEE remained negative (a CO₂ sink, −1610 ± 500 kg C ha⁻¹), because CO₂ lost during the drought was regained during the winter and spring. The site was also a net CO₂ sink during 2009 despite the colder than usual conditions (−2290 ± 500 kg C ha⁻¹). Including C imports and exports in addition to CO₂ exchange revealed that the site was a C sink in both years, with a NECB of 590 ± 560 kg C ha⁻¹ in 2008, and 900 ± 560 kg C ha⁻¹ in 2009. The C sequestration found in this study is in agreement with most other Northern Hemisphere EC studies of grazed pastures on mineral soils, but is not consistent with the large C losses reported for soils under dairy pastures throughout New Zealand. In the current study (like many other EC studies) the influence of climatic conditions and management practices on the annual C balance was only semi-quantitatively assessed. An extended period of EC measurements combined with modelling is required to more accurately quantify the effect of different climatic conditions on the annual C balance, and the influence of different management practices needs to be quantified using specifically designed studies (such as paired EC towers), so that practices which minimise C losses and maximise C sequestration can be identified.