Grazed grassland management regimes can have various effects on soil fauna. For example, effects on earthworms can be negative through compaction induced by grazing animals, or positive mediated by increases in sward productivity and cattle dung pats providing a food source. Knowledge gaps exist in relation to the behaviour of different earthworm species i.e. their movement towards and aggregation under dung pats, the legacy effects of pats and the spatial area of recruitment. The present study addressed these knowledge gaps in field experiments, over 2 years, using natural and simulated dung pats on two permanent, intensively grazed pastures in Ireland. Dung pats strongly affected spatial earthworm distribution, with up to four times more earthworms aggregating beneath pats, than in the control locations away from pats. In these earthworm communities comprising 11 species, temporally different aggregation and dispersal patterns were observed, including absence of individual species from control locations, but no clear successional responses. Epigeic species in general, but also certain species of the anecic and endogeic groups were aggregating under dung. Sampling after complete dung pat disappearance (27 weeks after application) suggested an absence of a dung pat legacy effect on earthworm communities. Based on species distributions, the maximum size of the recruitment area from which earthworms moved to pats was estimated to be 3.8 m2 per dung pat. Since actual grazing over 6 weeks would result in the deposition of about 300 dung pats per ha, it is estimated that a surface area of 1140 m2 or about 11% of the total grazing area can be influenced by dung pats in a given grazing period. This study showed that the presence of dung pats in pastures creates temporary hot spots in spatial earthworm species distribution, which changes over time. The findings highlight the importance of considering dung pats, temporally and spatially, when sampling earthworms in grazed pastures. Published comparisons of grazed and cut grasslands probably reached incorrect conclusions by ignoring or deliberately avoiding dung pats. Furthermore, the observed intense aggregation of earthworms beneath dung pats suggests that earthworm functions need to be assessed separately at these hot spots.