The methane (CH 4 ) and nitrous oxide (N 2 O) exchange of a temperate mountain grassland near Neustift, Austria, was measured during 2010–2012 over a time period of 22 months using the eddy covariance method. Exchange rates of both compounds at the site were low, with 97% of all half-hourly CH 4 and N 2 O fluxes ranging between ±200 and ±50 ng m −2 s −1 , respectively. The meadow acted as a sink for both compounds during certain time periods, but was a clear source of CH 4 and N 2 O on an annual timescale. Therefore, both gases contributed to an increase of the global warming potential (GWP), effectively reducing the sink strength in terms of CO 2 equivalents of the investigated grassland site. In 2011, our best guess estimate showed a net greenhouse gas (GHG) sink of −32 g CO 2 equ. m −2 yr −1 for the meadow, whereby 55% of the CO 2 sink strength of −71 g CO 2 m −2 yr −1 was offset by CH 4 (N 2 O) emissions of 7 (32) g CO 2 equ. m −2 yr −1 . When all data were pooled, the ancillary parameters explained 27 (42)% of observed CH 4 (N 2 O) flux variability, and up to 62 (76)% on shorter timescales in-between management dates. In the case of N 2 O fluxes, we found the highest emissions at intermediate soil water contents and at soil temperatures close to 0 or above 14 °C. In comparison to CO 2 , H 2 O and energy fluxes, the interpretation of CH 4 and N 2 O exchange was challenging due to footprint heterogeneity regarding their sources and sinks, uncertainties regarding post-processing and quality control. Our results emphasize that CH 4 and N 2 O fluxes over supposedly well-aerated and moderately fertilized soils cannot be neglected when evaluating the GHG impact of temperate managed grasslands.