An experiment was performed in a roe deer Capreolus capreolus (Linnaeus, 1758) population to assess the effect of a large-scale removal of adult males on the intensity of marking behaviour. Rubs and scrapes were censused in two 150-ha areas, one experimental and one control, in southeastern Sweden in May 1991 and 1992. In the experimental area, 14 males (50% of total number, 11 > 3 yrs) were culled during mid-August 1991. Removed males were replaced by the following season. Despite this, the number of scrapes, but not the number of rubs, was significantly fewer in the experimental area in May 1992 as compared to the control. This reduction could be due to a late arrival of replacement males, which would explain why the reduction involved scrapes rather than rubs, as scraping activity starts several weeks before rubbing. To cull adult males between territorial seasons is thus not an effective way to reduce rubbing intensity, and hence damages, on young forest plantations.