Earthworms are important in determining the greenhouse gas (GHG) balance of soils. In laboratory studies they have been shown to increase emissions of the potent GHG nitrous oxide (N2O). Here we test whether these earthworm-induced N2O emissions also occur in the field. We quantified N2O emissions in managed grassland in two different seasons (spring and autumn), applying two different types of fertilizer (organic and artificial fertilizer) and under two earthworm densities (175 individuals and 350 individuals m(-2)) of the species Lumbricus rubellus (Hoffmeister). We found an increase in earthworm-induced N2O emissions of 286 and 394% in autumn for low and high earthworm densities (P = 0.044 and P = 0.007, respectively). There were no effects of earthworms on N2O emissions in spring. Fertilizer additions significantly increased cumulative N2O emissions and grass N content in spring and autumn. For grass N content interactions between earthworm addition and fertilizer type existed in both seasons. Our results suggest that the pathways through which earthworms affect N cycling (and thereby N2O emission) differ with weather conditions. We postulate that in spring the dry weather conditions overruled any earthworm effects, whereas in autumn earthworms mainly improved soil aeration and thereby increased both plant N uptake and diffusion of N2O to the atmosphere. While we showed the presence of earthworm-induced N2O emissions in managed grassland under field conditions for the first time, the nature and intensity of the earthworm effect in the field is conditional on soil physicochemical parameters and thereby on meteorological and seasonal dynamics. (C) 2013 Elsevier B.V. All rights reserved.