Grazed pastures and cultivated fields are significant sources of greenhouse gas (GHG) emissions, in particular N ₂ O emissions derived from fertilizer deposition and animal excreta. Net surface emissions rely on subsurface gas transfer controlled mainly by diffusion, expressed as the soil‐gas diffusivity ( D _p / D _o ). The value of D _p / D _o is a function of soil air‐filled porosity (ε) and gaseous phase tortuosity (Ʈ), both of which vary with soil physical properties including soil texture and structure. Agricultural soils are often structurally aggregated and characterized by two distinct regions (inter‐ and intra‐aggregated pores), however, such soils are subjected to frequent compaction and tillage resulting in alteration to structural arrangement. In this study, a comparative analysis between the Currie (1960) and Taylor (1949) methods was performed to provide a computational insight into selecting an appropriate method for calculating D _p / D _o in agricultural soils. Currie's (1960) method was chosen for further analysis of the soils in this study. Results show that the D _p / D _o in aggregated soil cannot be expressed using a simple linear, power law or combined linear and power law functions due to the presence of two‐region characteristics. A new “Two‐Region model” was developed to parameterize the D _p / D _o of aggregated soils, and tested against repacked samples from two Sri Lankan agricultural soils. This Two‐Region model clearly distinguished tortuosity effects on gas movement with respect to density and textural variations within and between aggregates and outperformed previous models. The fitting parameters (α ₁ , α ₂ , β ₁ and β ₂ ) varied correspondingly with soil density, and the weighting factor ( w ) clearly distinguished the boundary between the two regions (inter‐ and intra‐aggregates) of structured soils. The model developed will be of interest to those seeking to model the diffusion of GHG emissions and gas exchange between the atmosphere and soils.