New Zealand dairy farming systems use considerable maize silage as a supplemental feed, often producing the crop within the farm boundary by cropping a paddock for one or more years then rotating through other areas of the farm. While globally the effects of permanent maize (Zea mays L.) cropping on soil carbon (C) stocks have been studied, little is known about the impact short-duration periodic cropping for maize silage within rotationally grazed pastures has on soil C stocks. In this study, we used the net ecosystem carbon balance (NECB) approach to determine the carbon budget of two maize silage crops in subsequent years and contrasted this with the alternative of grazed pasture on the same farm. Eddy covariance measurements provided CO2 exchange data, which were coupled with measurements or estimates of all other carbon imports and exports of the two systems. We chose to isolate just the period of maize silage production with our measurements extending from the time of herbicide application before planting of the maize crop through to seedling emergence of pasture re-establishment following harvest. The NECB of the maize silage system identified a large loss of C from the ecosystem (–850 g C m−2 and –415 g C m−2 for years 1 and 2 respectively), while the pasture was near neutral in year 1 (11 g C m−2) and a C source in year 2 (–114 g C m−2). Aboveground biomass production from the maize system was three times greater than the pasture system, with >90 % harvested and removed. In contrast, a net of ∼60 % of the aboveground biomass was removed from the pasture system. The large difference in exported biomass, coupled to long establishment periods with no photosynthetic C inputs of the maize were key factors in the much larger C losses from the maize system relative to the pasture.