Accelerated soil erosion can result in substantial declines in soil fertility and has devastating environmental impacts. Consequently, understanding if rates of soil erosion are acceptable is of local and global importance. Herein we use empirical soil erosion observations collated into an open access geodatabase to identify the extent to which existing data and methodological approaches can be used to develop an empirically-derived understanding of soil erosion in the UK (by way of an example). The findings indicate that whilst mean erosion rates in the UK are low, relative to the rest of Europe for example, 16% of observations on arable land were greater than the supposedly tolerable rate of 1 t ha ⁻¹ yr ⁻¹ and maximum erosion rates were as high as 91.7 t ha ⁻¹ yr ⁻¹. However, the analysis highlights a skew in existing studies towards locations with a known erosion likelihood and methods that are biased towards single erosion pathways, rather than an all-inclusive study of erosion rates and processes. Accordingly, we suggest that future soil erosion research and policy must address these issues if an accurate assessment of soil erosion rates at the national-scale are to be established. The interactive geodatabase published alongside this paper offers a platform for the simultaneous development of soil erosion research, formulation of effective policy and better protection of soil resources.