Incubation experiments are the most common approach to measure methane (CH4) oxidation potential in soils from various ecosystems and land-use practices. However, the commonly used headspace CH4 injection into microcosms and the shaking of the soil slurry during incubation fully removes CH4 (soil-born) and O2 (air-born) gradients common in situ, and may also induce various errors and disturbances. As an alternative, we propose CH4 input into microcosm soils via a silicone tube located within the slurry. We hypothesized that (i) poor CH4 diffusion in slurry will be compensated by direct CH4 delivery into the slurry via a silicone tube and, consequently, (ii) shaking of microcosms can be substituted with the soil silicone tube CH4 injection.During a 29-day submerged paddy soil incubation, the highest net CH4 oxidation rate was 1.6 μg C g−1 dry soil h−1, measured between the 3rd and 7th day after injecting 13CH4 into the slurry via a silicone tube without shaking. This rate was 1.5–2.5 times faster than the respective CH4 oxidation after headspace injection without shaking (1st hypothesis supported). As expected, shaking accelerated CH4 oxidation regardless of injection methods by 3.2–3.7 times (most intensively on days 3–7) compared to headspace injection without shaking. Nonetheless, the rates were similar between silicone tube injection without shaking and headspace injection with shaking. This supports the hypothesized potential of silicone tubes to substitute the common shaking method (2nd hypothesis). Furthermore, shaking increased the incorporation of 13C from CH4 into soil organic matter and microbial biomass by 1.8–2.7 times compared with CH4 injection into tubes and the static control without tubes. This reflects an overestimation of CH4 oxidation due to shaking. We conclude that direct soil CH4 injection via silicone tubes is advantageous in incubation experiments because gas concentration gradients are maintained and thereby more realistically reflect natural soil conditions.