Grasslands cover around 25% of the global ice-free land surface, they are used predominantly for forage and livestock production and are considered to contribute significantly to soil carbon (C) sequestration. Recent investigations into using ‘nature-based solutions’ to limit warming to <2 °C suggest up to 25% of GHG mitigation might be achieved through changes to grassland management. In this study we evaluate pasture management interventions at the Rothamsted Research North Wyke Farm Platform, under commercial farming conditions, over two years and consider their impacts on net CO2 exchange. We investigate if our permanent pasture system (PP) is, in the short-term, a net sink for CO2 and whether reseeding this with deep-rooting, high-sugar grass (HS) or a mix of high-sugar grass and clover (HSC) might increase the net removal of atmospheric CO2. In general CO2 fluxes were less variable in 2018 than in 2017 while overall we found that net CO2 fluxes for the PP treatment changed from a sink in 2017 (−5.40 t CO2 ha−1 y−1) to a source in 2018 (6.17 t CO2 ha−1 y−1), resulting in an overall small source of 0.76 t CO2 ha−1 over the two years for this treatment. HS showed a similar trend, changing from a net sink in 2017 (−4.82 t CO2 ha−1 y−1) to a net source in 2018 (3.91 t CO2 ha−1 y−1) whilst the HSC field was a net source in both years (3.92 and 4.10 t CO2 ha−1 y−1, respectively). These results suggested that pasture type has an influence in the atmospheric CO2 balance and our regression modelling supported this conclusion, with pasture type and time of the year (and their interaction) being significant factors in predicting fluxes.