The impact of deliberate culling of fox populations has been much debated. Although a local (< 10 km2) impact has been accepted, previous authors have denied that culling has any impact on a larger scale because local losses are compensated through immigration. Rather, it has been claimed that at this scale fox density is determined by resources, mediated through social behaviour and breeding suppression. We determined the impact of culling on a regional scale (> 1000 km2), using data on culling (Heydon & Reynolds, 2000), fox density (Heydon, Reynolds & Short, 2000) and productivity. The three U.K. study regions (size 1238–2322 km2) were in mid-Wales (A), the east Midlands (B) and East Anglia (C). High productivity in regions A and C was associated with low density, high culling mortality and high overall mortality (all relative to region B), indicating that density was suppressed by culling. In region B (moderate) breeding suppression was associated with a higher density and lower cull than in regions A and C, implying that fox density was closer to the maximum sustainable by resources. We conclude that the impact of culling in different regions of Britain is variable, dependent on the regional prevalence, methods, and history of culling. However, it is clear that in a range of circumstances culling can substantially depress fox numbers, and that current fox densities reflect a history of culling. This conclusion is fundamental in considering the management of fox predation in farming and conservation contexts.