Accurate prediction of gas diffusivity (Dp/Do) and air permeability (ka) and their variations with air-filled porosity (ε) in soil is critical for simulating subsurface migration and emission of climate gases and organic vapors. Gas diffusivity and air permeability measurements from Danish soil profile data (total of 150 undisturbed soil samples) were used to investigate soil type and density effects on the gas transport parameters and for model development. The measurements were within a given range of matric potentials (−10 to −500 cm H2O) typically representing natural field conditions in subsurface soil. The data were regrouped into four categories based on compaction (total porosity Φ <0.4 or >0.4 m3 m−3) and soil texture (volume-based content of clay, silt, and organic matter <15 or >15%). The results suggested that soil compaction more than soil type was the major control on gas diffusivity and to some extent also on air permeability. We developed a density-corrected (D-C) Dp(ε)/Do model as a generalized form of a previous model for Dp/Do at −100 cm H2O of matric potential (Dp,100/Do). The D-C model performed well across soil types and density levels compared with existing models. Also, a power-law ka model with exponent 1.5 (derived from analogy with a previous gas diffusivity model) used in combination with the D-C approach for ka,100 (reference point) seemed promising for ka(ε) predictions, with good accuracy and minimum parameter requirements. Finally, the new D-C model concept for gas diffusivity was extended to bimodal (aggregated) media and performed well against data for uncompacted and compacted volcanic ash soil.