An accurate representation of soil organic matter (SOM) decomposition dynamics in large scale Earth system models is required for a better prediction of carbon storage and its fluxes. Often soil scientists simplify the dynamics of soil carbon decomposition by assuming soils as well-mixed homogeneous compartments of varying substrate quality and microorganisms. However, soils are naturally heterogeneous, which results in a non-uniform spatial distribution of substrate carbon as well as microorganisms inhabiting it. Here, we formulate a methodology to account for spatial heterogeneities in soil carbon cycling models and investigate the limitations of using traditional models that neglect spatial variability.