Soils are the largest terrestrial store of carbon (C), but agriculture has depleted around 133 Pg C from the upper 2 m of soil, contributing to increases in atmospheric CO2. Soil C sequestration can help offset these atmospheric increases and is determined by the balance between C inputs and C outputs. New Zealand’s temperate climate supports year-round grazing on approximately one-third of the country’s land area. The conversion of New Zealand’s native forests to pastures increased soil C stocks, yet the mechanisms behind this increase remain unclear. Grazing animals influence the C cycle through the consumption of C embodied within the feed they eat, with excreta returning C to the soil. Although C balances of New Zealand pastures suggest excreta is an important return of C to the soil, this has been poorly quantified. The objectives of this thesis were to determine if soil C stocks (0–0.6m) differed between paddock and fenceline areas, assuming the paddock received more dung input. To test this assumption, the spatial distribution of dung in relation to the fenceline was mapped, and dung loading was estimated. Stocks of particulate organic matter C (POM-C) and mineral-associated organic matter C (MAOM-C) were also determined in the top 0.1 m of soil to understand which forms of C contributed to any measured differences in total C stocks. Differences in Olsen P and pH were also determined. Paddock and fenceline sites (38 pairs) distributed across six farms in the Waikato region were sampled to a depth of 0.6 m. On average, soil C and N stocks were 10.3 t C ha−1 (P = .002) and 0.96 t N ha−1 (P = .002) higher in paddock areas when compared to fenceline areas. When excluding one paired site with a very poor match of mineral surface area, the differences decreased to 9.3 t C ha−1 (P = .004) and 0.86 t N ha−1 (P = .003). The greatest differences were observed in the 0–0.1 m depth increment. At the 0–0.1 m depth increment, POM-C was significantly greater in paddock samples and largely accounted for the observed difference in total soil C. This suggested that the observed C stock increase was in a more degradable fraction, implying that decreases in dung returns could result in rapid soil C loss. There was no relationship between the age of the fenceline and C stock differences, indicating that differences may have occurred rapidly to a new steady state. Mapping of 2000+ dug pats showed that paddocks received more dung, with only one pat being within the fenceline sampling zone. The average dung density was 0.2 pats m−2 and a dung-C loading rate of 1.1 t C ha−1 y−1 was calculated. The detected differences in C stocks due to varying dung input aligned with other studies on cattle manure’s influence on soil C, POM-C and MAOM-C stocks. Further research is needed to better understand the contribution of dung returns to soil C and its potential vulnerability to loss.