Soil organic carbon (SOC) is a vital component of the global carbon cycle, underpinning soil health, ecosystem resilience, and climate change mitigation. Grasslands dominate global agricultural land and are often managed, placing significant SOC stores within the domain of land managers. However, they remain underrepresented in international research and measurement, reporting and verification systems. Improving the accuracy and transparency of SOC estimates is increasingly critical as global national frameworks predominantly rely on a simple default Tier 1 methodology. Emerging policies and guidelines favour Tier 3 modelling approaches, which capture local variability and management impacts beyond the scope of Tier 1 methods. This thesis advances Tier 3 SOC modelling for mineral grasslands through the development and application of a transparent, flexible modelling framework. A review of SOC models highlighted three key knowledge gaps: (i) model formulation, (ii) model inputs, and (iii) model applicability at the farm scale. These were addressed across three research chapters, focusing on the RothC model. Key findings include: (i) Model implementation - A generalised modelling framework was demonstrated to replicate RothC behaviour when appropriately amended and provides a common platform to focus future research efforts. (ii) Model input sensitivity - A multi-model ensemble analysis showed that simpler carbon input and evapotranspiration (ET) methods (e.g., Miami model, Thornthwaite ET) can perform as well as complex alternatives in data-limited contexts. (iii) Model development - A modular Mineral Grassland (MG) model was created, capable of operating at field and farm scales and accommodating future model extensions. (iv) Model evaluation - When tested on 28 Irish sites, the MG model showed good performance within the 73-105 Mg C ha-1 range, suggesting internal biases linked to the constrained current input configuration. Overall, this research develops and demonstrates a modular spatial Tier 3 SOC modelling framework for mineral grasslands, offering a transparent, adaptable platform to support farm-scale SOC accounting and ongoing model refinement.