Soil loss on arable agricultural land is typically an order of magnitude higher than under undisturbed native vegetation. Although there have been several recent attempts to quantify these accelerated fluxes at the regional, continental and even global scale, all of these studies have focused on erosion by water and wind and no large scale assessment of the magnitude of tillage erosion has been made, despite growing recognition of its significance on agricultural land. Previous field scale simulations of tillage erosion severity have relied on use of high resolution topographic data to derive the measures of slope curvature needed to estimate tillage erosion rates. Here we present a method to derive the required measures of slope curvature from low resolution, but large scale, databases and use high resolution topographical datasets for several study areas in the UK to evaluate the reliability of the approach. On the basis of a tillage model and land-use databases, we estimate the mean gross tillage erosion rates for the part of Europe covered by the CORINE database (6·5% of global cropland) and we obtained an average of 3·3 Mg ha–1 y–1, which corresponds to a sediment flux of 0·35 Pg y–1. Water erosion rates derived for the same area are of a similar magnitude. This redistribution of soil within agricultural fields substantially accelerates soil profile truncation and sediment burial in specific landscape positions and has a strong impact on medium-term soil profile evolution. It is, therefore, clear that tillage erosion must be accounted for in regional assessments of sediment fluxes and in analyses that employ these in the analysis of land management strategies and biogeochemical cycles. Copyright © 2009 John Wiley & Sons, Ltd.