Methane oxidation fluxes were monitored with the closed chamber method in eight treatment plots on a semi-wet grassland site near Giessen, Germany. The management regimes differed in the amount of nitrogen (NH4NO3) fertilizer applied and in the height of the in-ground water table. No inhibition of CH4 oxidation occurred, regardless of the amount of annual N fertilizer applied. Instead, the mean CH4 consumption rates were correlated with the mean soil moisture of the plots. However, the correlation between daily soil water content and corresponding CH4 oxidation rate was always weak. During drought period (late summer) water stress was observed to restrict CH4 oxidation rates. The findings led to the question whether methane production with soil depth might modify the CH4 fluxes measured at the surface. Therefore, two new methods were applied: (1) soil air sampling with silicone probes; and (2) anaerobic incubations of soil cores to test for the methane production potential of the grassland soil. The probe measurements revealed that the CH4 sink capacity of a specific site was related to the vertical length of its CH4 oxidizing column, i.e. the depth of the CH4 producing horizon. Anaerobically incubated soil cores produced large amounts of CH4 comparable with tropical rice paddy soil. Under field conditions, heavy autumnal rain in 1998 led to a dramatic increase of soil CH4 concentrations upto 51 microliters l-1 at a depth of 5 cm. Nevertheless, no CH4 was released when soil surface CH4 fluxes were measured simultaneously. The results thus demonstrate the high CH4 oxidation potential of the thin aerobic topsoil horizon in a non-aquatic ecosystem.