The turnover of greenhouse gases in soils is largely mediated by the activity of microorganisms. In situ quantification of these processes is important to improve estimates on global budgets for many greenhouse gases, including CH4. In this study, we assessed the utility of a modified version of the gas push–pull test (GPPT) to derive in situ apparent first-order rate coefficients for atmospheric CH4 oxidation in near-surface soil. An ordinary GPPT consists of the controlled injection and subsequent extraction of reactants (i.e., CH4 and O2) and tracer gases (e.g., Ar, Ne, or He) into and out of the vadose zone at a location of interest. In the modified GPPT, injection and extraction takes place inside a cylinder previously placed in the soil and temporarily closed at the top. Using numerical simulations and tests in a laboratory sand tank, we found gas recovery to remain low (3–17%) and decrease with decreasing injection–extraction depth during ordinary GPPTs of∼1-h duration. Under similar test conditions, modified GPPTs resulted in high gas recovery (64–90%) of all gases used. This allowed the modified GPPTs to be prolonged up to 12 h in simulations and 6 h in laboratory experiments, while gas recovery still remained substantial (19–57%). The modified GPPT was successfully applied to quantify atmospheric CH4 oxidation in situ in a sandy soil in Zurich, Switzerland. Calculated first-order rate coefficients ranged from 0.7 to 1.6 h−1 and agreed with literature values and estimates derived from a diffusion–consumption model fitted to CH4 concentration profiles. Our modification extends the applicability of GPPTs to study microbially mediated gas turnover in near-surface soils.