The gas push−pull test (GPPT) is a single-well gas-tracer method to quantify in situ rates of CH4 oxidation in soils. To improve the design and interpretation of GPPT field experiments, gas component transport during GPPTs was examined in abiotic porous media over a range of water saturations (0.0 ≤ Sw ≤ 0.61). A series of GPPTs using He, Ne, and Ar as tracers for CH4 were performed at two injection/extraction gas flow rates (~200 and ~700 mL min−1) in a laboratory tank. Extraction phase breakthrough curves and mass recovery curves of the gaseous components became more similar at higher Sw as water in the pore space restricted diffusive gas-phase transport. Diffusional fractionation of the stable carbon isotopes of CH4 during the extraction period of GPPTs also decreased with increasing Sw (particularly when Sw > 0.42). Gas-component transport during GPPTs was numerically simulated using estimated hydraulic parameters for the porous media and no fitting of data for the GPPTs. Numerical simulations accurately predicted the relative decline of the gaseous components in the breakthrough curves, but slightly overestimated recoveries at low Sw (≤0.35) and underestimated recoveries at high Sw (≥0.49). Comparison of numerical simulations considering and not considering air−water partitioning indicated that removal of gaseous components through dissolution in pore water was not significant during GPPTs, even at Sw = 0.61. These data indicate that Ar is a good tracer for CH4 physical transport over the full range of Sw studied, whereas, at Sw > 0.61, any of the tracers could be used. Greater mass recovery at higher Sw raises the possibility to reduce gas flow rates, thereby extending GPPT times in environments such as tundra soils where low activity due to low temperatures may require longer test times to establish a quantifiable difference between reactant and tracer breakthrough curves.